chess explained - ia600907.us.archive.org

Chess Explained: The GrUnfeld

Valentin Bogdanov

With contributions by Viacheslav Eingorn

Translated by Serge Marudov

First published in the UK by Gambit Publications Ltd 2009

'Copyright © Valentin Bogdanov 2009 English-language translation© Serge Marudov 2009

The right of Valentin Bogdanov to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1 988.

All rights reserved. This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, re-sold, hired out or otherwise circulated in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser.

ISBN- 1 3 : 978- 1 -906454-06-7 ISBN - 10: 1 -906454-06-X

DISTRIBUTION: Worldwide (except USA): Central Books Ltd, 99 Wallis Rd, London E9 5LN, England. Tel +44 (0)20 8986 4854 Fax +44 (0)20 8533 582 1 . E-mail: [email protected]

Gambit Publications Ltd, 99 Wallis Rd, London E9 5LN, England. E-mail: info@ gambitbooks.com Website (regularly updated): www.gambitbooks.com

Edited by Graham Burgess Typeset by John Nunn Cover image by Wolff Morrow Printed by the MPG Books Group in the UK.

10 9 8 7 6 5 4 3 2 1

Gambit Publications Ltd Managing Director: Murray Chandler GM Chess Director: Dr John Nunn GM Editorial Director: Graham Burgess FM German Editor: Petra Nunn WFM Webmaster: Dr Helen Milligan WFM

Contents

Symbols Introduction

1 2 3 4 5 6 7 8

Fianchetto Gri.infeld: White Plays g3 Non-Standard Lines with cxd5 Classical Exchange Variation: White Plays 4Je2 Modern Exchange Variation: White Plays 4Jf3 Modern Exchange Variation with l:[bl Russian System: White Plays 'ii'b3 Lines with i.f4 or i.g5 Closed Systems: White Plays e3

List of Games Index of Variations

4 5

6 24 38 52 61 83 98

116

125 126

Symbols

+ check ++ double check # checkmate ! ! brilliant move

good move !? interesting move ?! dubious move ? bad move ?? blunder Ch championship 1 -0 the game ends in a win for White lf2-lh the game ends in a draw 0-1 the game ends in a win for Black (D) see next diagram

Introduction

During the first quarter of the 20th century, revolutionary ideas took hold in the chess world. The classical opening principles, defined by Steinitz and polished by Tarrasch, carne under fire from all sides. One of the innovations was the opening first regularly employed by one of the top players of those years, the Austrian grandmaster Ernst Grtinfeld, whose name is forever reflected in the title of the opening.

w

The most traditional way to reach this position is via 1 d4 ltJf6 2 c4 g6 3 ltJc3 d5. Strictly speaking, our opening also encompasses variations where White delays the move ltJc3 or even avoids it altogether, most notably in systems with g3. However, the initial moves by Black always combine preparations for the fianchetto of the Icing's bishop with the provocative . . . d5, tempting White into building a large pawn-centre. Black hopes that his pieces, once developed, will create threats against that centre, while limiting the freedom of the enemy pieces to choose their residences.

The reputation of an opening depends on at what level and how often the opening is employed. Our opening has found favour with many outstanding chess-players and has often been used in competitions of the highest level, including world championship matches, with Botvinnik, Smyslov and Fischer among its staunch supporters. It was Kasparov's main choice in two of his matches with Karpov, which led to its second peak of popularity. Things are a bit quieter these days, but the Grtinfeld retains a large fan base among today's grandmasters and a worthy place under the sun.

The opening has sprung many branches, but despite the great variety of these systems, they can be divided into two main groups according White's strategy: either he picks up the gauntlet and acquires a pawn-centre early on, or opts for natural development, temporarily postponing hostilities in the centre.

This series of books does not aspire to encyclopaedic depth; the goal is to explain the ideas behind the opening and its individual systems, with particular emphasis on the more popular and important variations in recent practice. Of course, different players may have different opinions on popularity.

The chosen system for dividing the material into chapters doesn't always follow the traditional approach, but the author believes it to be the most logical.

1 Fianchetto Grunfeld : White Plays g3

As in many Closed openings, the fianchetto of White's king's bishop is a popular idea here as well. The pros and cons of this choice are universal: positioned on the long diagonal in relative safety, the bishop makes a worthy contribution to the white army's potential. However, the fianchetto requires a tempo to be spent on the move g3, and if the bishop is exchanged, then the legacy will be weak light squares, as a rule in the immediate vicinity of the king. A significant benefit for White is the fact that the fianchetto grants him additional move-order options, and leaves Black in the dark longer. White will decide how to deploy ofthe rest ofhis forces a few moves later, according to Black's actions.

Black's replies can be divided into two groups: a) . . . d5 is prepared by a preliminary . . . c6, to retake with the pawn in case of an exchange. This is

the orthodox approach to the problem of the centre and is contrary to the main strategic idea of the Griinfeld (we should note that Botvinnik classified systems with g3 as avoiding the Griinfeld Defence). Nevertheless, it gives Black an extremely solid position.

b) White is allowed to seize the centre, so as to create the counterplay typical of this opening, with the help of piece pressure and undermining pawn-thrusts.

Game 1 (Malakhov-Mamedyarov) examines the lines where in reply to . . . c6 and . . . d5 White immediately exchanges on d5. This leads to a symmetrical pawn-structure, and White will seek to capitalize on his first-move advantage. The game is of a relatively quiet character.

In Game 2 (Ribli-Volokitin), White does not hurry with the exchange on d5. Instead he attempts to force Black to relinquish the central outpost by means of piece pressure.

The remaining two games feature lines without . . . c6, where Black invites White to set up a pawn-centre in more typical Griinfeld fashion.

White plays in reserved fashion in Game 3 (P.H.Nielsen-Carlsen), consolidating his centre and gradually preparing a further advance. He ends up with a lone pawn on d5, and Black resorts to tactical measures in an effort to neutralize White's plans.

The pawns advance immediately in Game 4 (Koneru-Areshchenko), and once more a structure arises with an isolated white pawn on d5. While this pawn is a potential target, it also cramps Black and may fix the weakness on e7, or become a powerful passed pawn. We witness a lively debate on the pros and cons of this important formation.

Game 1 [079]

Vladimir Malakhov - Shakhriyar Mamedyarov Spanish Team Ch, Sant Uuis 2005

1 lLlf3 lLlf6 2 c4 g6 3 g3 1L.g7 4 1L.g2 c6 Instead of seeking piece pressure on White's

centre, Black chooses a more classical approach

by securing a pawn presence on d5. With the standard move-order, I d4 li:lf6 2 c4 g6 3 g3, if Black wishes to adopt this method, it is already

FIANCHETTO GRONFEW: WHITE PLAYS g3 7

high time for 3 . . . c6. Should this move be postponed, White has the time to play e4 with a likely transposition to the King's Indian Defence: 3 . . . ..ig7 4 ..ig2 c6 5 e4 d6.

5 0-0 It is more common to postpone castling for a

few moves in order to use the tempo more productively in the battle for the initiative. In particular, that's how Karpov treated this variation in several games in his matches against Kasparov. After 5 d4 d5 6 cxd5 cxd5 7l"Llc3 0-0 8l"Lle5 Black acquiesced to 8 . . . e6, which White can view as an achievement, though Black retains a sufficient reserve of solidity.

5 ... 0-0 6 d4 d5 7 cxd5 If White wants to maintain the pawn tension,

he can choose between 7 l"Llbd2, 7 'ii'b3 and 7 b3. The natural 7l"Llc3 allows 7 . . . dxc4 and is not as popular.

7 ... cxd5 (D)

w

Now the queen's knight can be developed to its most natural square, but Black has obtained the same privilege. The position is absolutely symmetrical; there is an open line, which is a likely avenue to mass exchanges of major pieces. White retains the first-move advantage and can fight for the initiative, but there is little tension left and the game often quickly ends in a draw. Of course, if both sides wish to find a way to continue the struggle, they always can. The symmetry can be broken with an exchange of knights on c3 or c6 (that is, by White playing l"Lle5 and meeting . . . l"Llc6 with l"Llxc6, or by Black playing . . . l"Lle4 and, if permitted, making the equivalent exchange on c3).

8 l"Lle5 This move takes control of c6 and hampers

the symmetrical knight jump to e4. The natural

alternative 8 ltJc3 invites Black to decide whether to accept the appearance of a pawn on c6 or to try to create such a pawn in White's camp:

a) After 8 . . . l"Llc6 9l"Lle5, Black should avoid 9 . . . l"Llxe5?! 10 dxe5 l"Llg4 1 1 l"Llxd5 l"Llxe5 12 ..ig5 with strong pressure for White, so the main reply is 9 . . . e6 10 l"Llxc6 bxc6. It's doubtful that Black will succeed in carrying out . . . c5, so White keeps a slight edge.

b) 8 . . . l"Lle4 is preferred if the goal is to give White a pawn on c3. Now 9l"Llxe4 dxe4 10 l"Lle5 ( 10 l"Llg5 'it'xd4 1 1 'it'xd4 ..ixd4 1 2l"Llxe4 steers the game towards a draw, except that now it is Black who has the more active position) promises White little, and Black can choose between several good continuations: 10 . . . 1i'd5, I O . . . l"Lld7 and 10 .. .f6. White can allow the exchange on c3 by 9 e3 or 9l"Lld2 when, having an extra tempo, he can carry out c4 and move into a standard middlegame.

8 ••• l"Llg4!? (D) As already mentioned, 8 . . . e6 is quite accept

able. After 9l"Llc3 Black has a further choice between 9 . . . l"Llfd7, which is considered the most solid continuation, and 9 . . . l"Llc6, inviting departure from the symmetry of the pawn-structure and starting the debate about the pros and cons of a pawn on c6.

w

9 l"Llxg4 9 f4 also promises little: after 9 . . . l"Llxe5 1 0

fxe5 ( 1 0 dxe5, leading t o less explored positions, has also been seen) Black will sooner or later play the undermining . . . f6, returning to a symmetrical pawn-structure and further manoeuvring play.

9 ... ..ixg4 10 l"Llc3 There are positive and negative sides of the

bishop's position on g4. Black cannot defend

8 CHESS EXPLAINED: THE GRONFEW

the threatened d5-pawn with the natural . . . e6 as long as the bishop is on this square; with its line of retreat cut off, the bishop would be quickly lost. On the other hand, the bishop prevents White from playing e3 and helps with the counterattack on d4.

The immediate I 0 h3 is well answered with I O . . . i.d7, and if 1 1 i.xd5 then 1 1 . . .lDc6, when two white pawns are en prise and one of them will be lost, as trying to hold on to the loot with 1 2 i.xc6 i.xc6 is too risky; light squares are more important.

10 •.• lDc6 1 1 i.e3?! (D) The generally accepted continuation is 1 1

h3, when I l . . .i.d7 1 2 e3 e6 looks the most solid, but I l . . .i.e6 is quite possible.

B

With the text-move, White persists in his attempts to demonstrate the weakness of the bishop's position on g4, and defends his pawn while keeping that of his opponent under fire. It's not clear whether that amounts to much if the symmetry is maintained with 1 1 . . . .i.e6, but Black has a more energetic continuation at his disposal.

ll ... e5! Now Black will obtain an IQP on the d4-

square. The cons of this pawn don't always outweigh its pros. In minor-piece endgames it limits the g7-bishop and as a rule White has the advantage, but with major pieces present, the pawn offers Black a space advantage and highlights the weakness on e2.

12 lDxd5 It is hard to explain the motives behind the de

cision taken by White in Wang Yue - VachierLagrave, Lausanne 2006: 1 2 dxe5? d4 1 3 i.xc6 dxc3 14 i.xb7 :lb8 1 5 .i.e4 cxb2 1 6 :lbl 'ifxdl 17 l:lfxdl J.xe5 1 8 i.xa7 .l:lbd8; with the enemy

pawn on b2, what advantage can White hope for?

12 ••. exd4 It makes sense to gain a tempo and keep

more pieces on the board. 13 .i.cl 1 3 .i.f4 can be met by 1 3 . . . g5, taking the last

retreat -square away from the knight and provoking promising complications: 14 i.cl lle8 1 5 :ei :e5 1 6 h3 i.e6 1 7 e4 f5.

13 ... :es (D)

w

The weakness on e2 is already subjected to pressure and more than balances out the potential weakness of the pawn on d4. Black has the advantage in development and space, while his pieces can be more easily relocated for play on the flanks. Nevertheless, at the moment Black's advantage is minimal.

14 l:te1 l:tc8 (D)

w

15 h3 Initiating the transfer of the knight to the

blockading square d3. The immediate 1 5 iDf4 is met by the unpleasant 1 5 . . . lDb4, while the prophylactic 1 5 a3 runs into 1 5 . . . d3 1 6 'ifxd3 ltJd4 and White must pick the lesser evil and


part with the exchange after 1 7 �e3. Therefore he drives the bishop away first, and then re-stricts the enemy knight. w

15 ... �f5 16 a3 'ii'd7 17 'iith2 �e4 The exchange of light-squared bishops is

part of Black's plan of strengthening the position - the enemy king's cover is weakened and his light squares become vulnerable to the manoeuvres by the black pieces.

18 lt::lf4 �xg2 19 'iitxg2 lt::le5 20 lt::ld3 (D)

B

20 ••• lt::lc4 This manoeuvre does not bring any addi

tional benefits. The c3-knight bolsters White's defences and

Black's priority should be to exchange it off; the question is how. 20 . . . 'ifb5 or 20 . . . 'ii'd5+ looks good, gradually improving the piece placement, but note the attempt to win a pawn immediately doesn't achieve its goal because of the weakness of the back rank: 20 . . . 'ii'd5+ 21 'iitgl lt::lxd3 22 'ii'xd3 'ii'e4 23 �f4 %:tc2 24 'i¥xe4 .lhe4 25 %:tac l .

21 �d2 White, of course, resists the temptation to

play b3, as the resulting hole on c3 would portend a catastrophe once the black knight is transferred to that square.

21 ••• 'ii'b5 22 %:tb1 h5 (D) This pawn raid is typical and useful: allowing

.. . h4 is undesirable, and the symmetric counter weakens the squares around the white king even more.

23 h4 %:te4 White has no chronic weaknesses, but nei

ther has he a hint of counterplay nor a chance to provoke a crisis, while his opponent can test various piece configurations at his leisure. Playing this kind of position is psychologically

unpleasant. The e4-rook fulfils several functions - it exerts pressure down the half-open e-file, it overprotects the pawn on d4, and can participate in the attack on the white king. Moreover, there is no way for White to drive the rook away.

24 �f4 %:teeS 25 'ifc2 lt::lb6 26 %:tbd1 It is more logical to occupy the open file.

While White does leave the c 1 -square free for the bishop, it could do just as fine on g5; should Black drive it away with .. .f6, the resulting weakening could come in handy in the future.

26 ••• lt::lds 21 �cl b6 2s 'iitn ?! One possible way to strengthen Black's posi

tion involves the transfer of the bishop to d6, when . . . f5-f4 or the knight sacrifice on e3 will have to be taken into account. Therefore White protects e2 in order to offer a knight exchange. However, the king's position on f l is precarious, which gives Black various tactical chances.

w

28 ••. 'ii'd7 (D)

29 lt::lf4 There is nothing better. 29 ••• lt::lxf4 Black could have taken advantage of the

unfortunate position of the enemy king with 29 . . . �h6 30 lt::lxd5 (30 lt::lg2 loses to 30 ... 'ifh3 3 1

10 CHESS EXPLAINED: THE GRONFELD

l:tg 1llJf6 32 f3 .i.xc1 33 l:txc1 l:tg4!) 30 . . . �xd5. Then 3 1 i..xh6 is bad because of 3 l . . .l:txe2 with mate on h 1 , while 3 1 'it>g 1 i.xcl 32 l:txc 1 d3 brings White little joy, and on 31 e3 the most incisive is 3 1 . . . l:txe3 32 .l:.xe3 .i.xe3 33 .i.xe3 �f3.

30 .i.xf4 .i.f6 The blow at h4 cannot be averted. 31 �d2 3 1 e3 Io_ses to 3 1 . . . �h3+ 32 'it>gl dxe3, when

any recapture is met by the exchange sacrifice on e3, sweeping away all the kingside pawns.

3 1 'it>g2 .i.xh4 32 e3 .i.f6 33 f3 l:.4e7 34 e4, closing the e-file, looks more stubborn, but here too Black's extra pawn and the white Icing's weakened cover should bring Black victory.

31. .• .i.xh4 32 gxh4 g5 (D) It seems that White underestimated this re

source. 33 f3 This is tantamount to resignation, but after

33 hxg5 �h3+ 34 'it>gl �g4+ 35 .i.g3 h4 36

w

�d3 Black has a pleasant choice between 36 . . . l:t8e5 and 36 . . . l:tf4 threatening 37 . . . l:te3.

33 ... l:txf4 34 'it>g2 �f5 35 e4 dxe3 36 l:txe3 l:txe3 37 �xe3 l:txh4 38 l:td8+ 'it>h7 39 �c3

White even manages to weave a mating-net, but Black is first to land the blow, of course.

39 ... �h3+ 40 'it>f2 �h2+ 41 'it>fl �h1+ 42 'it>f2 l:th2+ 43 �e3 �g1 + 44 'it>e4 l:th4+ 45 'it>eS l:tf4 46 b4 "i'g3 47 �d3+ l:tf5++ 0-1

Game 2 [0781

Zoltan Ribli - Andrei Volokitin Slovenian Team Ch, Ce/je 2004

1 liJf3 g6 2 d4llJf6 3 g3 i..g7 4 .i.g2 d5 5 0-0 c6 6 liJbd2

White's move-order avoids the 6 c4 dxc4 (D) variation that is considered rather pleasant for Black.

w

Now there are several gambit continuations, such as 7llJc3, 7llJbd2 and 7llJa3, when White obtains at least some compensation for the pawn; whether it is sufficient is a moot question.

The pawn can be regained with 7 a4, but that requires time expenditure and weakens squares on the queenside. Black gets comfortable de-velopment; for example, 7 . . . llJa6 8 llJa3 0-0 9 llJxc4 .i.e6 10 b3 .i.d5 with . . . c5 to follow.

7llJe5 deserves a separate mention. It fails to regain the pawn because Black can counterattack against d4 by 7 . . . llJg4, when 8 f4 0-0 9 e3 llJxe5 10 fxe5 c5 1 1 llJa3 llJc6 is more than pleasant for Black. The game Bologan-Volokitin, Sarajevo 2005 is of interest: 8 llJxc4 �xd4 9 �c2 0-0 1 0 llJba3 �f6 I I h3 llJe5 1 2 llJa5. The compensation is hardly adequate, although if tense irrational play was White's wish, then it has been fulfilled.

6 ... 0-0 7 c4 Unlike in the previous game, White refrains

from the exchange on d5 and continues the struggle for pawn superiority in the centre, preparing e4 or hoping to force Black to take on c4.

This pawn move does not have to be linked with the move liJd2, and there are some moveorder issues here, as we have already seen in the


previous note. To avoid confusion, we shall take the position arising after 1 d4 lDf6 2 c4 g6 3 g3 i.g7 4 i.g2 0-0 5 lDf3 c6 6 0-0 d5 (D) as the starting point for our discussion of these alternatives:

w

a) 7 l2Jc3 posts the knight more actively, but after 7 . . . dxc4 the game takes on a gambit character. Although such gambits have a right to exist, Black has a fully adequate, and indeed promising, game. 8 l2Je5 is well met with either 8 ... i.e6 or 8 . . . l2Jg4, while the prophylactic 8 a4, aimed against the possible . . . b5, as a rule keeps the pawn safe, at the cost of a loss of time and the weakening of the squares on the b-file.

b) 7 b3 (D) protects the c4-pawn at once, while also clearing a spot on the long diagonal for the bishop.

B

White is still planning to prepare the e4 advance. Black can attempt to capitalize on his temporarily greater control of the long diagonal by 7 . . . dxc4 8 bxc4 c5 9 i.b2 cxd4 10 lDxd4. He has then spoiled White's pawn-structure, but at the cost of activating the enemy pieces. The other method is relatively tame: fighting for

piece control of e4. Here is a possible scenario: first the black knight leaps to e4, and waits for White to develop his queen' s bishop. Then after an exchange of knights the c8-bishop relocates to the newly vacated e4-square to neutralize its counterpart on g2.

c) 7 lDe5 is similar to the previous game, but with the pawns on c4 and c6 intact. In this scenario 7 . . . e6 is too disrespectful of the c8-bishop, but 7 . . . l2Jg4 is quite good. A certain slowness of White's development can even justify Black in surrendering the centre with 7 ... i.e6 or 7 ... dxc4 for the sake of rapid piece deployment followed by a counterattack: 7 . . . i.e6 8 cxd5 i.xd5 9 i.h3 c5, Speelman-Kasparov, World Cup, Reykjavik 1988.

d) 7 'ili'b3 (D).

B

This is the most popular move. The plan is to increase the pressure on the d5-pawn while restraining the c8-bishop by the attack upon b7. Of the many possible replies we shall note three:

d l ) 7 . . . a5 is an attempt to harass the white queen and seize the initiative on the queenside. After 8 lDc3 a4 9 'ili'b4 dxc4 10 'ilfxc4 'ili'a5 , despite ceding the centre, Black acquires sufficient freedom to develop his pieces comfortably.

d2) 7 . . . dxc4 is an immediate surrender of the centre in order to take advantage of the early development of the white queen: Black plans to mobilize his forces with gain of tempo.

d3) 7 . . . 'ili'b6 brings about a characteristic situation, seen in many openings, where the queens sit in opposition to one another. It is often hard for both players to judge whether to exchange or to allow the opponent to make the exchange. On the one hand, doubled b-pawns can prove somewhat weak, but the half-open a-file can be useful too. The decision tends to


depend on subtle positional nuances, which need to be continually re-evaluated as long as the queens remain on b3 and b6. After 8 ltJc3 l:td8 (several other continuations have been tested, but today the main discussion is centred around this reply), White can force the matter by 9 c5 'ii'xb3 10 axb3 ltJa6 I I iLf4, with chances for both sides. White aims to push e4; Black, correspondingly, . . . e5 .

e) Finally, 7 ltJbd2 transposes to the position we are about the examine in the main game.

Thus we now return to 7 c4 (D):

B

This approach is fairly popular. The knight supports the preparations for e4 equally well from d2 while defending c4 at the same time. Of course, on d2 it is posted less actively than on c3 while blocking the bishop's diagonal. However, it is rare for a move to have only pluses and no minuses.

7 ... a5 Anticipating the fianchetto of White's queen's

bishop, Black demonstrates his readiness to initiate play on the queenside in order to seize space and open the a-file. That does not mean that e4 receives less attention; on the contrary, control of this square is the main goal.

Black's other main options are to play more directly in the centre with 7 . . . iLf5 or 7 . . . ltJe4.

When choosing 7 . . . iLf5, Black hopes, after 8 b3 ltJe4 9 iLb2, to achieve the exchange of his light-squared bishop either via e4 (9 . . . ltJxd2 10 'it'xd2 iLe4) or via h3 (9 . . . ltJd7 10 ltJh4 ltJxd2 I I 'it'xd2 i.e6 12 e4 dxe4 1 3 iLxe4 iLh3 14 l:tfe1 liJf6 1 5 iLg2, Ki.Georgiev-Shirov, Russian Team Ch, Sochi 2007).

These same positions frequently arise even after the immediate 7 . . . ltJe4, though 8 b3 gives Black the additional possibility of 8 . . . c5 9 iLb2

ltJxd2 10 'it'xd2 dxc4 1 1 bxc4 cxd4 I 2 liJxd4, when White's pieces are active but his pawns are broken. 8 e3 is also possible, already threatening to take on e4; after 8 . . . ltJxd2 9 ltJxd2 White controls e4, but his knight has been decentralized and his development has not advanced.

We now return to 7 ... a5 (D):

w

8 b3 8 ltJe5 is worth mentioning here as well, with

the standard 8 . . . ltJg4 as an adequate reply. 8 ... a4 (D) 8 . . . iLf5 and 8 .. . ltJe4 are played in this posi-

tion as well, which often comes down to a transposition of moves. However, the plan with . . . c5 makes little sense at this point since it doesn't harmonize well with . . . aS and means not just the loss of a tempo but also weakened squares along the b-file.

w

9 iLb2 This is the most natural move. At the mo

ment, the bishop's scope is limited by the d4-pawn, but White hopes that the pawn will acquire mobility whe� e4 is achieved. It's also possible to develop the bishop to a3, temporarily


closing the a-file. In that case, Black generally endeavours to keep the position closed, placing the e-pawn on e6. This is preceded by . . . .tg4, since leaving the queen' s bishop stuck in its own camp would be too passive. This leads to an exchange of this bishop for a white knight, whereupon White can hope in the long run to exploit his bishop-pair. It's possible for Black to offer an exchange of the dark-squared bishops with ... .tf8; while the offer is hard to decline, a result of the exchange is some weak squares around the black king. Achievements of this sort are of limited benefit to White, though, as Black' s resources are quite sufficient to maintain approximate equality.

White can also play 9 bxa4, opening files on the queenside. 9 b4 is also possible, with the opposite idea, viz. to avoid opening the position. However, after 9 . . . dxc4 Black gets comfortable squares for his pieces, and weaknesses (the b4-pawn or light squares in case of a3) now appear in the white camp as well.

9 .•• lt:Je4 (D) Again 9 . . . .tf5 is more common here, but the

text-move is a tempting idea: Black unmasks his dark-squared bishop, places c3 under control, and forces White to take . . . a3 into account.

w

10 lt:Jxe4 10 bxa4 leads to simplifications after 10 .. . lt:Jc5

1 1 a5 .l:txa5 1 2 lt:Jb3 lt:Jxb3 13 axb3 .l:txa 1 , while 10 'ifc 1 fails to prevent 10 . . . a3: 1 1 .txa3 lt:Jxd2 12 lt:Jxd2 .txd4 (Vukic-Ruck, Banja Luka 2007). The text-move changes the pawn-structure and sharpens the coming struggle.

10 ••. dxe4 U lt:Jel f5 12 lt:Jc2 White's plans include undermining the en

emy pawn-chain, but 12 f3 allows an immediate counterattack by 1 2 . . . c5 1 3 lt:Jc2 lt:Jc6.

Therefore it is a sensible idea to bolster the d4-pawn first.

12 ... lt:Jd7 13 lt:Je3 13 e3 has been played, to which 13 . . . lt:Jc5

looks like a good reply. 13 ... lt:Jc5 (D) 13 . . . e5 leads to unclear complications after

14 d5 f4 1 5 lt:Jc2 f3 1 6 exf3 exf3 1 7 .txf3 .l:txf3 1 8 dxc6 bxc6 19 'ifxf3 e4 20 'ifxe4 .txb2, while Black may even play 13 . . . f4 at once.

w

White has no concrete threats, but what is he to do? The queen must simultaneously protect b3 and d4. 14 b4 is bad because of 14 . . . a3; that leaves us with . . . .

14 :bt axb3 15 axb3 (D)

B

15 ... .l:ta2 Supporting the assault on b2 and making

White face the same problems again. 16 .tal White has regrouped, and the c5-knight is en

prise. If Black now fails to capitalize on the temporary activity of his pieces, he will have to sound a retreat, and it is not clear how well the a2-rook will manage it.

16 ... .l:txal (D)


w

Consistent, at least; retreating the knight would render Black's preceding play dubious, while now he gets a pawn for the exchange and sentences White's light-squared bishop to life imprisonment.

17 :xa1 .txd4 18 :aS?! What did the b3-pawn do to deserve this

treatment? After the natural 1 8 :bl .txe3 1 9 fxe3 'ifxd l 20 :fxdl .te6 (D) an unclear position would appear.

w

The g2-bishop is a typical example of a piece locked out of the game; its habitat is restricted to the fl -h3 diagonal (plus the h 1 -square ). Even pushing g4 doesn't change the situation - the f5-pawn has a lot of defenders. It follows that Black has the upper hand on the queenside and risks little. On the other hand, it's not easy for him to win by playing on one flank alone, and dynamic action in the centre runs the risk of activating the enemy bishop. A sample variation: 21 b4 lt:Ja4 22 :bct1;g7 23 c5lt:Jb2 24 :d4 :as.

The active sortie by the rook, chosen by White in the game, is a mistake.

18 .•. -txeJ 19 fxe3 Here the exchange on d8 by 1 9 'ifxd8 would

have hindered the development of the c8-bishop.

In that case Black has no good reason to grab the pawn by interpolating 1 9 . . . ..txf2+, because this only helps White activate his forces and gives the prisoner on g2 some hope for the future.

19 ..• 'ii'xd1 20 .:txd1 lt:Jxb3 21 g4?! White hastens to get this move in before

Black plays . . . h5, in order to keep the h-pawns on the board and have at least some sort of a target to attack. However, it is better to focus attention on the b7-pawn, by 2 1 .:ta3 lt:Jc5 22 :bi :ds 23 :aS, preventing the bishop's development. Black must then seek an effective regrouping plan.

21. .• .te6 (D)

w

22 :dd8 White strives to keep a rook on the 8th rank

in order to attack the h-pawn. However, this is the quickest way to lose, as the black pawns are faster. It is more stubborn to try to hang on to c4, temporarily tying the enemy bishop to the defence of f5 by 22 .:txf8+ 1;xf8 23 gxf5 gxf5 24 .th3, or keeping both rooks on the board with 22 :a7.

22 .•• :xd8 23 l:txd8+ 1;f7 24 :hs <tlg7 25 :e81;f6 (D)

w

• •n• 'l••• . . ...

........ � ... � B B E -'�w/""

. . ..... ·�···��

• • • n •7//////, -·· d •

• -�·.t� • • • m - . . �

FIANCHEITO GRUNFELD: WHITE PLAYS g3 15

26 h4?! Here, too, the counterattack on f5 promised

more - 26 gxf5 gxf5 27 l:tb8 lt::lc5 28 i.h3 i.xc4?! 29l::tf8+, although Black has a won position, of course.

26 ••. �xc4 27 g5+ �e6 28 h5 gxh5 There is no reason to allow the pawn on to

h6; a passed g-pawn is easier to contain. 29 .l:th8 b5 30 ..th3 b4 31 l::tf8 lt::ld2 32

i.xf5+ �e5 33 ..txh7 b3 34 .l:tb8 i.b5 0- 1

Game 3 [076]

Peter Heine Nielsen - Magnus Carlsen Bundesliga 2004/5

Our focus in this game moves on to lines where White's cxd5 is met by . . . lt::lxd5, with Black foregoing the move ... c6 in favour of more active methods.

As in other branches of the Fianchetto Griinfeld, there are many possible move-orders by which these lines can be reached, with a number of subtle differences in the possibilities that they offer to both sides. The choice of each player depends on what opening variations he is willing to allow or wishes to avoid, which in tum is determined by personal taste and the level of preparation as well as by the study of the corresponding factors in the opponent's games.

1 lLlf3 lLlf6 2 c4 g6 3 g3 i.g7 4 i.g2 0-0 5 0-0 d5

With this move, the game moves closer to Griinfeld territory. White's delay in playing d4 gives Black options of . . . c5, with a Symmetrical English.

6 cxd5 We should also consider the continuation 6

d4 dxc4 (D).

w

This capture is obviously a possibility whenever White plays c4 after Black has already

made the move . . . d5. Conversely, if White wishes to avoid these lines, he needs to play an earlier c4 and meet . . . d5 with cxd5.

We should note immediately that Black isn't surrendering the centre in the expectation of keeping the extra pawn; instead he hopes to take advantage of the time White will spend regaining it. That said, 7 lt::lbd2 or 7 lt::lc3 implies a readiness for gambit-like play on White's part, and these moves should be compared with lines we saw in the previous game. 7 lt::le5 offers Black the choice between 7 . . . c5 and an attempt to hang on to the extra soldier for a while by 7 . . . lt::le8 8 lLla3 tLld6.

The main continuation is 7 lt::la3, which is usually met with 7 . . . c3 (D).

w

The question arises - why change White's pawn-structure in such a way as to bolster the d4-pawn? There are several reasons: • Now the g7-bishop hinders the advance of

the d-pawn. • If the bishop's opposite number is developed

to b2, it will be staring at the back of its own pawn-chain and for some time to come won't see much action.

16 CHESS EXPlAINED: THE GRONFELD

• Black may be able to create an outside passed pawn in the endgame. Here is the typical treatment of this posi

tion: 8 bxc3 c5 9 e3 ltJc6 10 11i'e2 .i.f5 1 1 l:.d 1 1!i'b6 1 2 .i.b2 l:.fd8 1 3 ltJc4 1!i'a6 14 i.fl l:.ac8, and each side has his trumps, Ki.GeorgievAreshchenko, Coventry 2005.

6 ••. ltJxd5 7 ltJc3 Let's consider yet another line - 7 d4 c5 (D).

w

This immediate counterattack in the centre can be met in several ways:

a) 8 dxc5 is the most natural. After 8 . . . ltJa6 the plan is to pose Black some problems with the manoeuvre 9 ltJg5 e6 (Black's difficulties persist after 9 . . . ltJdb4 as well) 10 ltJe4.

b) 8 e4 is another popular option. 8 . . . ltJf6 allows White a pleasant game after 9 e5 ltJd5 I 0 dxc5 ltJa6 I I a3 ltJxc5 12 b4 ltJe6 13 'ii'b3, while the other knight retreat leads to play more typical of this opening: 8 . . . ltJb6 9 d5 e6 10 ltJc3 exd5 1 1 exd5 .

c) Another common position, often reached via different move-orders, can arise after 8 ltJc3 cxd4 9 ltJxd4 ltJxc3 10 bxc3. White is trying to prove that the activity of his pieces is more important than his ruined queenside. The course of events can be depicted with the pawn sacrifice in the game Aronian-Grishchuk, French Team Ch 2005: I 0 . . . 1!i'c7 I I .l:!.bl 'ifxc3 I 2 ltJb5 'ii'a5 13 i.d2 1!i'd8 14 i.e3 'ii'a5 1 5 'ife 1 .

7 ... ltJc6 8 d4 ltJb6 Now we reach a key decision-point, where

White must choose whether to advance or support his d4-pawn.

Once more, there are some important moveorder issues to highlight. Suppose Black doesn't like the idea of White being able to advance his d-pawn so freely. Could he avoid this situation?

With certain sequences, yes. Let's consider the position arising after 1 d4 ltJf6 2 c4 g6 3 ltJf3 i.g7 4 g3 d5 5 cxd5 ltJxd5 6 i.g2 ltJb6 (by voluntarily withdrawing from the centre in good time and saving a tempo on castling, Black creates pressure on the d4-pawn) 7 0-0 ltJc6 (D).

w

This is the same as the position in the main game except that Black hasn't castled and White hasn't played ltJc3. The pawn has come under attack before it has had a chance to advance. White is faced with a choice:

a) He can defend the pawn by 8 e3. If Black is happy just to have avoided the line with a quick d5 advance, he can transpose to standard lines by 8 . . . 0-0 (see the course of the main game), or take an independent course by 8 . . . e5 9 ltJc3 exd4 (9 ... 0-0 transposes to the note to Black's 9th move below) 10 exd4 0-0 (as long as the check by the rook from e1 is possible, no one is tempted by the capture) I I i.g5 with a fairly complicated version of the IQP formation.

b) 8 ltJc3 is an interesting pawn sacrifice. Black can decline by 8 . . . 0-0, transposing to the main game, but obviously the critical reply is 8 . . . ltJxd4 9 ltJxd4 'i¥xd4 (rarely seen is 9 .. . i.xd4, agreeing to the immediate loss of castling; in Drozdovsky-Korchnoi, Odessa 2007 White had strong pressure after 10 ltJb5 i.e5 I I 'ifxd8+ �xd8 1 2 l:.d l + ltJd7 13 f4 i.g7 14 i.e3 a6 1 5 ltJa7 c 6 1 6 l:.ac I ). White can regain the pawn at once with the relatively forcing I 0 ltJb5 'i¥c4 I I 'ifb3 0-0 1 2 'ifxc4 ltJxc4 13 ltJxc7 l:.b8, but with accurate play Black has good equalizing chances. A real gambit arises after I 0 'ifxd4 i.xd4 I I ltJb5 i.e5 12 i.f4 i.xf4 13 gxf4 �d8 14 l:.fd I + ltJd7 with chances for both sides.

White can avoid these lines, if he wishes, by picking a different move-order, as, for example,


in the main game (as always, at the cost of giving Black extra options). Therefore we now re-turn to the position after 8 . . . lDb6 (D): B

w

This is one of the most popular positions in the Fianchetto Grtinfeld. The white d-pawn can either advance at once, or wait to be forced to do so by the move . . . e5. The decision is not clear-cut, as the black pawn on e5 constitutes both a foothold in the centre and a blockage of the g7-bishop's diagonal.

9 e3 This generally implies the latter approach, as

Black will soon need to play . . . e5 to generate counterplay.

The alternative is of course 9 d5 lDa5. Now: a) 10 tiJd4 doesn't make much sense as the

knight will be hanging after IO . . . c6. b) 10 'iVc2 is a relatively quiet plan; White

hopes to gain a tempo or two and get ahead in development after IO . . . c6 II dxc6 lDxc6 12 �d I 'iVe8, but the resulting symmetrical pawnstructure limits the scope of the subsequent play.

c) 10 e4 c6 leads to a struggle more typical of the Grtinfeld. Now White can post his bishop on f4 at once, or first provoke a weakness on the kingside by II �g5 h6 12 �f4. A sample line runs 12 . . . cxd5 1 3 exd5 lDac4 14 'iVe2 (the sharp 14 b3 �xc3 15 �c I requires more practical tests). Now Black should avoid the initially tempting line l 4 . . . lDxb2?! 15 'iVxb2 lDa4 16 lDxa4 �xb2 17 lDxb2 as three pieces are worth more than the queen in the middlegame - even with an extra pawn thrown in.

We now return to 9 e3 (D). The d4-pawn prefers to be asked to advance

with . . . e5 ; after all, the g7-bishop's diagonal is blocked then and it's easier to acquire passed status.

9 ... l:.e8 For a long time the main continuation was

9 . . . e5 10 d5. Then: a) Normally the knight moves to a5 or e7,

but we should also mention 10 . . . e4, when it is hard to see anything better than 1 1 dxc6 'iVxd 1 12 l:xd l exf3 1 3 �xf3 bxc6. Black's active pieces offset the defects in the pawn-structure. Nevertheless, White succeeded in casting doubt on this idea in Tkachev-Sutovsky, European Ch, Dresden 2007: 14 �d2 lDc4 15 lDa4 �b8 16 �c3 �xc3 17 bxc3; Black's activity is gradually suppressed, and the weaknesses remain.

b) IO . . . ltJa5 presents White with a nice chance to obtain a small but solid advantage with II e4 c6 12 �g5 f6 1 3 �e3 cxd5 14 �xb6 'iVxb6 15 lDxd5 'iVd8 1 6 �c I lDc6 17 'iVb3.

c) The retreat IO ... lDe7 has been losing its appeal: II e4 �g4 12 h3 �xf3 13 'iVxf3 c6 14 l:td l cxd5 15 exd5 lDf5 16 d6 and the problems remain.

The fact that 9 . . . e5 does not appear to be very pleasant for Black leads one to consider alternatives. For the time being, White's position doesn't look particularly aggressive, so Black can afford a quiet preparatory move, challenging White to find a similarly useful move. One possibility is 9 . . . a5, hoping for 10 b3 e5, when 1 1 d5 (II �a3 is better) is met by l l . . . e4. However, the benefits of . . . a5 are not so clear, so the main continuation nowadays is to use the rook to reinforce the advance of the e-pawn.

We therefore return to 9 . . . l:te8 (D): 10 d5 What has changed since last move that made

White consider the formerly rejected pawn advance now to be favourable? Firstly, the move e3 is useful, in that it affords the d4-square additional protection. This means that a knight on


w

that square won't be hanging, and this in tum hinders the freeing . . . c5 or .. . e5 advances. Secondly, what other continuation would be useful? The black rook has prophylactically evaded the possible tempo-winning attack by the bishop from a3, so 10 b3 can be met by IO . . . e5, forcing the exchange on e5 with comfortable equality.

There remains 10 .l:te1 as an alternative to the text-move. Black will then sometimes keep insisting that White should disclose his plan first, usually with IO . . . a5 or IO . . . ..td7. However, IO . . . e5 is more common, when the following sharp line is topical: 1 1 d5lt:'la5 1 2 e4 c6 1 3 b3 cxd5 14 exd5 e4 1 5lt:'lxe4 .l:txe4 ( 1 5 . . . ..txa1 at once is bad because of 16 ..tg5, while after 1 5 . . . lt:'lxd5 1 6 ..tg5 f6 1 7 ..td2lt:'lc6 1 8 .l:tc l the white pieces are noticeably the more active) 1 6 .l:txe4 ..txa1 1 7 ..tg5 ..tf6 18 ..txf6 'ili'xf6 1 9 .l:te8+ �g7 20 'ili'e I ..tg4 21 .l:txa8lt:'lxa8 22lt:'le5 and 23 b4, winning back the piece and retaining an extra pawn, Podgaets-Nedilko, Odessa 2007.

10 ••• lt:'la5 lllt:'ld4 This move prevents the advance of the c- and

the e-pawns and blocks the enemy bishop's diagonal, allowing b3. Moreover, . . . c6 now requires preparation.

ll ••• ..td7 (D) Can the undermining advance . . . c6 be pre

vented? 12 e4 White decides simply to get on with his own

play in the centre. Other moves: a) 12 a4, threatening to push through all the

way to a6 and knock out the c-pawn's support, is interesting. Indeed, the immediate 1 2 . . . c5 leads to difficulties after 1 3 dxc6 lt:'lxc6 14 a5 and 1 5 a6, but a good reply is 1 2 . . . .l:tc8, to meet 1 3 b4 with 1 3 . . . c5.

w

b) If White plays 1 2 b4 first, then after 1 2 . . . lt:'lac4 1 3 a4 a5 14 b5 'ili'c8 1 5 .l:te1 Black can secure adequate counterplay with either 15 . . . ..th3 or 15 . . . e5 1 6lt:'lb3 (the simplest reply to 1 6 dxe6 is 1 6 . . . ..txe6, when the bishop can-not be taken because of the threat of the capture on c3) 16 . . . c6.

c) 1 2 b3 leads to a quiet game. After the immediate 1 2 ... c6 1 3 dxc6 (the standard push 1 3 e4 i s hampered by the c3-knight's unprotected state - 1 3 . . . cxd5 14 exd5 e6) 1 3. .. lt:'lxc6 14 lt:'lxc6 ..txc6 1 5 ..txc6 bxc6 White has a symbolic advantage. In the absence of immediate threats there is enough time for the preparatory 12 . . . .l:tc8 as well.

12 ••• c6 13 .l:tel cxd5 14 exd5 (D)

B

This is one of the key positions of the Fianchetto Griinfeld. The d5-pawn, while isolated and potentially weak, for the moment gives White a space advantage and fixes the weakness on e7. We have seen a similar formation, with the colours reversed, in Game 1 ; however, here there is a full set of pieces on the board, and not all of them can be said to be placed well. Black's top priority is the transfer of a knight to the blockading square d6.

FIANCHETTO GRONFELD: WHITE PLAYS g3 19

14 ••• .l:lc8 14 .. .'�Jac4 can also be played at once. In

most cases this comes down to a mere transposition of moves after 1 5 b3 li:ld6 16 .i.f4 .l:lc8, but then White is given a chance to develop the bishop to b2, and here it is Black who acquires additional options.

15 .i.f4 With the obvious intention of offering an ex

change on e5. 15 .•• li:lbc4 Why this knight? In the light of the follow

ing play it doesn't matter which, and the b6-square is freed for the queen just in case. Black is clearly aiming for the subsequent sharp line, and the knight is not headed for d6, when the other one would look silly stranded on a5 . Other moves:

a) 1 5 ... li:lac4 1 6 b3li:ld6 is more natural and solid, followed by the transfer of this knight to f5, while its colleague goes to d6 via c8.

b) 15 ... h6 is interesting, and provoked sharp and unclear complications after 16 .i.e5 .i.xe5 1 7 l:r.xe5 li:lbc4 1 8 .l:le1 li:lxb2 19 'i!Vd2 'i!Vb6 20

li:le6 in Grishchuk-Timofeev, Russian Ch, Moscow 2004.

16 b3 (D)

B

16 ••• li:lb2 Black was perhaps under the impression that

he was about to force a draw. 17 'ti'd2li:lac4 18 bxc4li:lxc4 It seems as if, tied to the defence of d4, the

queen cannot break away from the irritating knight's pursuit. However, Nielsen's home analysis disclosed something that Carlsen had missed: first the weakening .. .f5 is forced, then the queen relocates to b3 to take advantage of the newly opened diagonal.

19 'i!Vd3 tt'lb2 20 'i!Ve3 li:lc4 21 'i!Ve4 f5 22 'i!Vd3li:lb2 23 'i!Vd2li:lc4 24 'i!Vd1li:lb2 25 'ii'b3 .i.xd4 26 d6+ (D)

B

The long forced line has ended and White demonstrates the results of his analysis.

26 .•. e6 Avoiding 26 . . . li:lc4? 27 .i.d5+, 26 . . .<tig7? 27

'i!Vxb2 .i.xc3 28 i.e5+ and 26 . . . �h8? 27 'i!Vxb2 .l:lxc3 28 .i.e5+.

27 tt'ld5 .i.a4 This move appears forced. The capture on d5

is, naturally, out of the question, while White threatened the prosaic li:le7+.

28 'i!Vxb7 (D) About half a year later White preferred 28

'iVb4 in Borovikov-Areshchenko, Ukrainian Ch, Rivne 2005; Black rejected 28 . . . .i.xf2+ 29 �h l .i.xe1 30 'i!Vxb2 l:r.c2 3 1 'i!Vxb7 .i.c3 32 li:lxc3 'ili'f6 with an unclear game, but his continuation 28 . . . .l:lc4 29 'i!Vd2 �f7 turned out to be inferior.

B

28 ... l2Jd3 29 l2Je7+ �h8? This loses. The struggle remains compli

cated after 29 ... .l:lxe7 30 dxe7 'i!Ve8. Now all White has to do is reach the long diagonal and it's checkmate.


30 .te3 l:tc4 30 . . . .txal fails to save Black: 3 1 l:hal l:tb8

32 'ili'a6 .tb5 33 'ili'a3 and the numerous threats cannot be met.

31 'ili'a6 ltJxel After 3 l .. . .txe3 32 l:xe3 Black will lose at

least a piece. 32 'ili'xc4 .txa1 33 .td4+ e5 34 'ili'e6 1-0

Game 4 [072]

Humpy Koneru - Alexander Areshchenko Lausanne 2006

1 d4 ltJf6 2 c4 g6 3 g3 .tg7 4 .tg2 d5 5 cxd5 ltJxdS 6 e4 (D)

B

White takes the maximum advantage of the opportunity to seize the central squares. The result of this strategy depends on whether he can maintain the superiority in the centre or transform this advantage into another kind.

On the ever-present question of move-orders, we should note that the approach used by White here is not available if he has already played either ltJf3 or ltJc3, the former being particularly common in many players' repertoires.

6 ••• ltJb6 The knight has two reasonable squares to

move to, so we should also examine 6 . . . ltJb4. The d4-pawn is attacked three times (taking the tactical trick . . . ltJc2+ into account), so it must advance either at once or after a preliminary a3, shooing the knight away; alternatively White can sacrifice the pawn. Thus:

a) The unclear sacrifice 7 ltJe2 .txd4 8 0-0 .tg7 9 'ili'a4+ gives White the freer game and better development. Naturally, Black is not keen to exchange off his g7-bishop.

b) 7 a3 drives away the knight. Although it wins a tempo, it forces the knight to a more active post. After 7 . . . ltJ4c6 8 d5 ltJd4 Black gets

counterplay whether White plays 9 ltJc3 c5 10 ltJge2 e5, when the position acquires a closed character, or temporarily allows a pin with 9 ltJe2 .tg4 10 ltJbc3 c6 1 1 0-0 ltJxe2+ (the sharp l l . . .liJf3+ 1 2 �hl h5 is also interesting) 1 2 ltJxe2 0-0.

c) 7 d5 is more common, intending to drive the knight still farther away, but that is not easy to accomplish after 7 . . . c6, as shown by the line 8 a3 'ili'a5 9 ltJc3 cxd5 10 .tf4 (the attempt to reinforce d4 with 10 .te3 is met with 10 . . . ltJc2+ 1 1 'ili'xc2 d4) 10 ... ltJ4c6 1 1 exd5 ltJd4. A cunning idea is 8 ltJe2 and only after 8 . . . cxd5 to play 9 a3, when after 9 . . . ltJ4c6, 10 exd5 does not allow the knight into d4, whereas the alternative post on e5 is less favourable. Instead, 9 . . . 'ili'a5 is risky due to 10 0-0 dxe4 1 1 .td2.

7 ltJe2 (D) The pawn is under attack, and as long as it

can be undermined from either side, pushing it forward is illogical, as is allowing a pin with 7 ltJf3 .tg4, unless the idea is to secure the two bishops for the future by 8 d5 c6 9 h3.

B

7 ... c5 Another example of the counterplay charac

teristic of the Griinfeld can be seen after 7 . . . 0-0 8 0-0 ltJc6 9 d5 l2Ja5 10 ltJbc3 c6.


8 d5 Now White will normally obtain a passed

pawn. The attempt to get the more active pieceplay by 8 dxc5 'ii'xd l + 9 �xdl lba4 10 lbbc3 lbxc5 1 1 .i.e3 lbba6 1 2 'iti>c2 cuts both ways -the white king doesn't feel safe enough in the centre.

8 •.• e6 9 0-0 0-0 10 lbec3 The queen's knight can also be stationed

here by 10 lbbc3. The king's knight can use the f4-square. The text-move leaves the queen's knight free to choose a route to c4, b5 or e4.

10 ••• exd5 11 exd5 (D)

B

A typical pawn-formation has appeared on the board; similar ones can be found in other openings too. What factors might favour White here? • a central passed pawn, penetrating into the

enemy camp, which secures a slight advantage in space;

• control of the c6-square, the black knight's natural post;

• most of Black's pieces are on the queenside, so his king is a little short of defenders;

• there are prospects of an attack on the c5-pawn. We should also consider the features rele

vant to Black's counter-chances: • for the moment, the advance of the passed

pawn is not secured; • its support is limited to pieces only, so it is li

able to become weak; • concrete play aimed at creating a forward

post on d3; • queenside pawn-majority.

ll ... lba6 This appears the most natural: from this

square there are two routes to d3, as well as the

possibility of joining in the attack on the d5-pawn; the blockade of the passed pawn is not necessary at the moment. In general in this line, White can afford to meet the bishop sortie . . . .i.f5 with g4 - he is not afraid of opening lines on the kingside, where he has more forces, while Black is forced to lose time.

12 lLld2 The knight is bound for e4 to work on two

targets - c5 and f6. With the development incomplete, 12 d6?!

looks dubious, but 12 a4 and 1 2 lLla3 are sometimes seen, both of which are prophylactic in nature. Then 12 ... .i.f5 can be met with the already-mentioned idea 1 3 g4 or with 1 3 .i.e3, striving for the quickest mobilization, when Black can prepare the exchange of the lightsquared bishops with 1 3 . . . 'ii'd7.

12 ••• lbb4 A pawn-grab by 12 . . . .i.xc3? 1 3 bxc3 lbxd5

is suicidal; after 14 lbe4 there is a dangerous draught blowing through the dark squares in the black camp.

The usual continuation is 12 . .. .i.f5 1 3 tt'lde4 'ii'd7 and now 14 .i.g5 .i.xe4 1 5 lbxe4 f5, but there is also the tactical shot 14 .i.h6 with the threat of 1 5 .i.xg7 �xg7 1 6 lbxc5 and 17 'ii'd4+.

13 lbde4 c4 (D) Here too, the capture of the pawn does not

look tempting: 1 3 . . . lb4xd5 14 lbxd5 lbxd5 15 lbxc5.

w

Black's queen's knight is obviously planning to park itself on d3. White must clearly fight against this, but first provokes a weakening of the a2-g8 diagonal and significantly limits the g7-bishop's scope.

14 .i.g5 f6 15 .i.e3 lbd3 16 lbc5


White would like to keep the pawns on the board, but defending b2 allows Black to take control of c5 with the queen on e7. After the pawn exchange Black acquires play down the c-file. The next few moves are virtually forced.

16 ••• lt:lxb2 17 'ii'd2 lt:ld3 18 lt:lxd3 cxd3 19 'ii'xd3 .i.fS 20 'iWbS .l:.c8

An earlier game saw a less successful continuation: 20 . . . 'ii'd7 2 1 .i.xb6 axb6 22 'ii'xb6 l:lfc8 23 .l:.ac 1 .i.h6 24 lt:le4 .i.xe4 25 .l:.xc8+ .l:.xc8 26 .i.xe4, Gevorgian-Baikov, Moscow 1 995. Despite the opposite-coloured bishops, the exposed king and the pawn deficit leaves Black few hopes for salvation.

21 .l:.acl .l:.f7 This multi-purpose move covers the 7th rank,

prepares to double the rooks and vacates a square whereby the bishop can be transferred to a promising diagonal.

22 .l:.fd1 lt:lc4 (D)

w

This position is critical for the evaluation of the plan initiated by Black's 12th move. It seems that the capture on a7 (23 .i.xa7) could now have cast doubt on it. Such a decision demanded deep calculation, for which White, perhaps, didn't have enough time.

23 .i.cS?! b6 24 'iWxc4 l:txcS 25 'ii'b3 How should we evaluate the resulting posi

tion? In the middlegame White has a strong passed pawn in the centre, though at the moment it is not easy to set it in motion, whereas Black's 2-to- 1 pawn-majority on the queenside will be more important in the endgame. The white knight has a path to c6 (or, should a favourable opportunity arise, to e6) where its exchange, and consequently the relocation of the passed pawn, would usually be unfavourable for Black, though it is possible to intercept the

knight earlier. Black has two powerful bishops, and the dark-squared one is particularly strong as it has no opponent; the fact that its scope is limited by the pawn on f6 is not a concern -there are other diagonals too.

2S ••• .i.f8 The preliminary 25 . . . .i.h6, driving the rook

away, is interesting idea. After 26 f4, which is a useful move for Black to provoke, the plan is to initiate counterplay with 26 . . . g5.

26 lt:lbS 26 d6, steering the king to g7, is unclear. It is

true that the g2-bishop comes to life, but the pawn may be simply lost.

26 ••• .l:.xcl 27 l:xcl .i.cS (D)

w

The bishop has taken up an excellent post, and White must be careful not to allow an attack on f2.

28 'ii'c3 As it would not have been easy to decide on

the line 28 d6 �g7 29 .l:el a6 30 .i.d5 .l:.f8 (30 . . . .l:.d7? loses to 3 1 .i.c6 axb5 32 .i.xd7) 3 1 .l:.e7+ �h8 3 2 ..ig8 g5, White prefers a quiet game and prepares the transfer of the knight to d4.

28 ••• �g7 29 lt:ld4 .i.d7 Black could probably hold the position with

out difficulty by exchanging on d4 with a subsequent . . . .l:c7, but that meant switching to defence without counterplay; he hopes for more.

30 h4 A characteristic decision. Having found that

neither 30 lt:lc6 - Black would not exchange -nor 30 lt:le6+ promises anything tangible, White plans to open a second front: she hopes to induce weaknesses with the help of the pawn-thrust or to include another combat unit in the fight.

30 ••. .l:.e7 31 h5

FIANCHEITO GRUNFELD: WHITE PLAYS g3 23

Here 3 1 l2Jc6?! forces the exchange, but the counterattack against f2 hands Black the initia-tive: 3 1 . . . �xc6 32 dxc6 .:.e2 33 .:.n .:.xa2. w

3I. .. .:.es It is more sensible to get rid of the pesky

pawn with 3 1 . . .gxh5, acquiescing to a certain weakening of the kingside structure.

32 �f3 'fie7? (D)

w

However, this careless move leads at once to a difficult position. Black fails to take into account the fact that the e2-square is no longer available.

33 l2Jc6 �xc6 34 dxc6 What did the relocation of the passed pawn

to c6 change? It seems at first that there is still nothing to support its march across the dark squares, given the opposite-coloured bishops. First, there is only one such square now, which provides White with additional tactical possibilities; second, the farther the pawn advances, the larger space advantage it secures, which translates into increased piece mobility with regard to play on two fronts.

34 .•. 'fic7 The queen sets its sights on the g3-pawn;

now 35 . . . .:.e3 is threatened. 35 'iti>g2 White is alert, and there at once follows a

losing reply: 35 ..• .:.e7? (D) However, Black has no truly satisfactory

continuation.

36 h6+ 'iti>f7 37 �d5+ 'it>e8 38 'fixf6 Black's greatest misfortune is not so much

the loss of the pawn as the terrible position of the king.

38 ••. 'fie5 Black is forced to seek the transition into a

hopeless endgame. The game could probably be finished quicker with the queens on the board, but the continuation in the game is the simplest. The further play requires no comments.

39 �f7 + 'iti>d8 40 .:.d 1 + 'iti>c8 41 Vi' xeS :xeS 42 .:.d7 .:.e7 43 .:.xe7 �xe7 44 �g8 �f8 45 �xh7 �xh6 46 �xg6 'iti>c7 47 �e4 bS 48 f4 aS 49 'iti>f3 b4 50 g4 a4 51 gS �f8 52 fS b3 53 axb3 a3 54 �bl 'it>xc6 55 'iti>e4 'it>cS 56 f6 'iti>b4 57 g6 'iti>xb3 58 g7 �xg7 59 fxg7 'it>b2 60 g8'fi 1-0

The systems with g3 continue to claim a sizeable section in the general body of theory of the Griinfeld Defence. Black has considerable freedom to dictate the strategic direction of the play, most notably with the choice between the solid yet passive systems with . . . c6 and . . . d5, or the debate on the typical Griinfeld theme of the pawn-centre and the struggle against it. However, especially in the latter case, White can choose between a variety of variations of differing characters.

While isolated improvements for each side are constantly introduced, the overall evaluation gives Black no particular cause for alarm.

2 Non-Standard Lines with cxd5

What do we mean by the standard version of the Exchange Grtinfeld? As already mentioned in the notes to Game 3, Black benefits in many cases from the transfer of the b2-pawn to c3, which is a feature of I d4 �f6 2 c4 g6 3 �c3 d5 4 cxd5 �xd5 5 e4 �xc3 6 bxc3, which can be considered the main line of the whole opening. The exchange of knights, the cause of this transfer, is another plus - White enjoys an advantage in space under the cover of the pawn-centre, and one of the principles of chess strategy states that in many positions exchanges benefit the side with less space.

In this chapter we discuss lines where White exchanges pawns on d5, but in some manner avoids the change of structure resulting from bxc3. This can be accomplished by the knight being absent from c3, or by preparing to recapture with a piece on c3.

Game 5 (Laznicka-Ponomariov) is dedicated to the move 3 f3, which has been known for a long time but didn't enjoy a great reputation until recent years, when it became fairly popular. White prepares e4 without putting the knight on c3. However, this deprives the knight of the natural developing square and costs a tempo that could be used on development. Moreover, as a possible practical drawback it commits White to the Siimisch Variation if it turns out that Black prefers to play the King's Indian Defence.

In Game 6 (Lastin-Yandemirov) the move e4 is prepared with 5 i.d2, in order to recapture with the bishop, avoiding a structural weakness while also neutralizing the powerful enemy bishop on g7. This is a quiet, indeed almost sterile, line sometimes used when White wants to play with an absolute minimum of risk. Black's best chance for a full-blooded battle is to avoid the exchange on c3 by playing .. . �b6. While this retreat would be almost unthinkable in the standard Exchange Variation (see the note to Black's 5th move in Game 5), here the fact that White has played i.d2 is more or less a waste of a tempo.

Finally, Game 7 (Toulzac-Bogdanov) examines Nadanian's fantastically paradoxical idea 5 �a4. This move caused something of a sensation when it was introduced, but once one has overcome the initial shock caused by this move's blatant eccentricity, the logic behind it is quite clear. Nevertheless, it is no surprise that Black has found some reasonable antidotes.

Game 5 [085]

Viktor Laznicka - Ruslan Ponomariov Karlovy Vary 2007

1 d4 �f6 2 c4 g6 3 f3 (D) This move dates back to Alekhine's games,

from the earliest years of Grtinfeld history. White prepares e4 without playing �c3, but allows Black to pick one of several reputable systems from other openings - Modem Benoni, Benko Gambit and the already-mentioned Siimisch King's Indian. Evidently, this move is

most expedient when you know what to expect of your opponent - or else if you have a repertoire into which the move f3 fits seamlessly.

3 ... d5 4 cxd5 �xd5 5 e4 �b6 We should compare this position to the simi

lar one that arises after I d4 �f6 2 c4 g6 3 �c3 d5 4 cxd5 tiJxd5 5 e4 �b6?! . This knight retreat (as well as the move-order I d4 �f6 2 c4 g6 3

NON-STANDARD LINES WITH cxd5 25

B

lbf3 d5?! 4 cxd5 lbxd5 5 e4 lbb6 6 h3; however, as long as White is not ready to play e4, the reply . . . d5 can wait) has been condemned by theory, with 6 h3, forestalling the pin on g4, recognized as the best continuation. Subsequent events can develop in several ways, but Black falls short of drumming up sufficient play, as in the example line 6 . . . il..g7 7 lbf3 0-0 8 il..e2 tt::lc6 9 il..e3 f5. The situation in the game is more promising for Black since the white knight is deprived of the f3-square and White's position is looser, although the general pattern of counterplay is similar (but more effective!).

6 lbc3 .tg7 7 .teJ 0-0 (D) The immediate 7 ... lbc6 presents White with

a choice between 8 il..b5 0-0 9 lbge2, maintaining a flexible pawn-centre, and 8 d5 lbe5 9 f4 (9 il..d4 0-0 1 0 f4 lbg4 most often transposes) 9 . . . lbg4 10 il..b5+ (or 1 0 il..d4) 1 0 . . . .td7 1 1 il..d4. These lines are not thought critical for Black, but why give the opponent additional possibilities? Putting off castling won't save tempi.

w

8 'ili'd2 In the line 8 f4 lbc6 9 d5 tt::la5 10 il..d4 il..g4

(or 10 ... e5 at once) 1 1 .te2 e5 Black's advantage

in development counterbalances White's central preponderance.

8 •.. lbc6 The central counterblow 8 . . . e5 9 d5 c6 is also

possible. White usually embarks at once on the attack with 10 h4, a typical advance in this sort of position. Then Black can allow the h-pawn to advance, but this probably isn't an ideal solution; e.g., 10 . . . cxd5 1 1 exd5 lb8d7 12 h5 lbf6 1 3 hxg6 fxg6 14 d6 il..e6 1 5 lbh3 gave White the better chances in Laznicka-Areshchenko, Moscow 2007. The pawn can be halted with 10 . . . h5, a move used by Shirov to win in his match against Kramnik, although the result was by no means a direct consequence of the opening. Later Ivanchuk suggested that White should play vigorously on the kingside, not balking at a pawn sacrifice, and the scales started tipping in White's favour: 10 . . . cxd5 1 1 exd5 h5 12 g4 hxg4 1 3 h5 il..f5 14 hxg6 fxg6 1 5 d6 lb8d7 16 lbd5 .:tf7 1 7 il..g5 lbf6 1 8 lbxb6 'i!i'xb6 1 9 .tc4, Moiseenko-Timofeev, European Clubs Cup, Saint Vincent 2005.

9 0-0-0 (D)

B

This is the basic position of this variation. It has much in common with analogous positions in other openings, such as the Sicilian Dragon, the Sarnisch King's Indian and the related system in the Pirc Defence. White threatens a daring attack with a small force - h4-h5xg6, il..h6, exchange bishops and checkmate along the hfile. In this instance there is also a distinction: the presence of a mobile pawn-centre (not a pawn-wedge as in the King's Indian). The pawn-centre is particularly attractive when the pawns stand next to each other on the same rank, with each threatening to advance; that limits the enemy pieces' options to a higher

26 CHESS EXPLAINED: THE GRUNFELD

degree. This is why White protected the d4-pawn instead of making the d5 advance. For that purpose the king had to be sent to the queenside, and opposite-side castling often heralds violent mutual attacks. Also notable is the transfer of the knight from f6 to b6, where it doesn't take part in the defence but reinforces the assault group instead. Building the centre caused a certain lag in White's development, although both the fl -bishop and the h 1 -rook can prove usefully placed where they stand. The moment is ripe for a counterblow by Black in the centre.

9 ... f5 (D) This move has gained recognition in recent

years. For a long time the move 9 . . . e5 occupied centre stage. While blocking the bishop, it secures an influential square for the knight. After 10 d5 lt:Jd4 (the sacrifice - or blunder - of a piece can hardly be correct: 10 . . . lt:Ja5? 1 1 .ixb6 axb6 12 b4, Rustemov-Illescas, Dos Hermanas 2005) 1 1 f4 c5 1 2 fxe5 .ig4 13 :e1 .ixe5 14 h3 'i!kh4 1 5 .id3 .l:lac8 there ensues a sharp, double-edged struggle.

w

At the first glance it is not clear what Black's threats are, except that after an exchange on e4 he will obtain the g4-square. Actually, the point is that with the opening of the f-file, the flbishop comes under attack and the capture on d4 therefore becomes a possibility. If White replies with 10 .ie2 then 10 . . . e5 gains in strength.

10 h4 At present, of all the continuations only two

enjoy wide acceptance. One is the text-move, and the other is 1 0 e5. This move concedes the d5-square but as compensation locks the g7-bishop out of the game. Here we should mention the game Motylev-Svidler, Wijk aan Zee

2007: 10 . . . lt:Jb4 1 1 lt:Jh3 ( 1 1 h4 is played more often) 1 l . . ..ie6 12 �b1 'i!kd7 13 lt:Jf4 .if7 14 a3 aS - a standard piece sacrifice for the sake of the initiative, which White failed to neutralize in that game.

If it proves impossible for White to maintain his broad pawn-centre, he needs to find the best way to transform this advantage into positional achievements of a different kind - a development advantage, more active pieces, etc. One such method is the attempt to use the time Black has to spend destroying the centre to launch a rapid attack on his king. That is the aim of the text-move.

10 ... fxe4 11 h5 (D) We have already mentioned what could hap

pen after 1 1 fxe4, while in the event of 1 1 lt:Jxe4 .ie6 1 2 b3 a5 the question arises: who is attacking whom?

B

ll ... gxh5 Black secures a diagonal for his bishop,

where it can take part both in the defence and the counterattack.

It is obvious that 1 1 . . . exf3 helps White's development, and is not worth the extra pawn.

The immediate counter in the centre 1 1 . . .e5 has lost its former popularity. After 12 d5 lt:Jd4 13 hxg6 hxg6 14 fxe4 ( 14 .ih6 is met by the jab 14 . . . e3, slowing down the development of the attack) White has the better chances.

12 :xhS If the capture on h5 is to be postponed, then

the most obvious idea is the exchange of the dark-squared bishops. 12 .ih6 at once leaves the d4-pawn without protection, so the line 1 2 d5 lt:Je5 13 .ih6 has been tested. However, Black's counterplay demonstrated in the game Aleskerov-Sanikidze, World Junior Ch, Istanbul 2005

. NON-STANDARD LINES WITH cxd5 27

retired this continuation, perhaps only for the time being: 1 3 . . . tDec4 (this must be played at once, before the bishops are exchanged and the d4-square becomes available to the queen) 14 'ii'g5 .l:If7 1 5 .l:Ih4 'ii'd6 1 6 i..xg7 .l:Ixg7 1 7 'ii'xh5 i..d7 1 8 .l:Ixe4 'it'b4 (once the queen makes it to this square, it can create very dangerous threats) 1 9 .l:Id2 'ii'c5 20 i..xc4 tDxc4 2 1 .l:Idd4 lDd6 22 .l:If4 lDf5 and the exchange had to be given up, after all.

12 •.. i..f5 (D)

w

13 .l:Ig5 This move appears the most flexible: as the

attack along the h-file has slowed down, the rook is switched to exert pressure along the gfile, at the same time evading a possible attack. Now the threat is fxe4.

Instead, 1 3 i..h6 is again met with 1 3 . . . 'ii'xd4 and the exchange-grab 14 'ii'xd4 i..xd4 1 5 i..xf8 is bad because after 1 5 . . . .l:Ixf8, the threats of . . . i..e3+ and . . . i..xg1 cannot both be met. The preliminary 1 3 d5 runs into the zwischenzug 1 3 . . . i..g6, and in the line 14 .l:Ih3 lt:le5 1 5 fxe4 lt:\ec4 1 6 i..xc4 lt:lxc4 1 7 'ii'e2 lt:lxe3 1 8 'ii'xe3 c6 the white king is as uncomfortable as his black counterpart.

13 ... i..g6 Taking the pawn would speed up the knight's

arrival onto the battlefield; this is also an indirect defence against 14 fxe4, as the fl -bishop would again be exposed to attack.

14 i..e2 White is forced to waste a tempo, and Black

carries out the standard counter in the centre. 14 ••• e5 15 d5 l2Jd4 Here Black can already snatch a pawn with

1 5 . . . exf3, but after 16 lt:lxf3 lt:ld4, White can play 1 7 lt:\h4, after which his initiative fully

compensates for the loss; on the other hand, nothing is achieved with 1 7 lt:lxe5 lt:\xe2+ 1 8 'it'xe2 .l:Ie8 1 9 lt:\xg6 'ii'xg5 20 i..xg5 .l:Ixe2 2 1 lt:\xe2 hxg6 and the endgame i s level.

16 fxe4 c6 17 dxc6 17 lt:lf3 has been tested as well, but Black

achieves a fully adequate game: 17 . . . i..f6 1 8 .l:Ig3 lt:\xe2+ 1 9 'ii'xe2 cxd5, although now instead of 20 exd5 .l:Ic8 (Moiseenko-Li Chao, Moscow 2007), White could have tried 20 lt:\xd5 or 20 i..xb6.

17 .•. lt:lxc6 The recapture with the pawn would open one

file opposite the enemy king but close the other one, whereas the c-file looks more promising with the current piece configuration. Now the queen exchange would result in approximate equality, but White hopes for more.

18 'ii'e1 'ii'f6 (D)

w

The evaluation of this position is critical for the assessment of the plan of counterplay initiated with . . . f5.

While for the moment neither of the black bishops takes an active part in the events, they cement Black's defence and secure a powerful outpost for the knight on d4; an exchange on this square would blast open the centre for the bishops and lead to a powerful attack on the white king. Even without the bishops, Black has the advantage in the number of combat units on the queenside. It is not clear if the white rook is attacking or is stuck on g5.

The conclusion must be that Black's chances are at least no worse. Now for the first time there is a real threat of . . . i..h6 (earlier it would have been refuted with .l:Ixg6+ ), which should have been met by retreating the rook or moving the bishop to h5 immediately.


19 Vg3?! The queen's kingside sortie is a risky deci

sion. 19 •.• ltJd4 20 i..h5 The situation urgently called for the prophy

lactic 20 �bl . After the text-move, the rook finds itself on an unprotected square, which gives Black an important tempo.

20 •.• i..xh5 21 J:r.xh5 l:.ac8 22 �b1? After 22 i..h6 Vg6 23 Vxg6 hxg6 24 i..xg7

gxh5 25 i..xf8 .l:txf8 White still can offer resistance, although Black already has a palpable advantage in the endgame.

22 ••. l:.xc3 (D) Somewhat unexpected, yet quite logical -

the white king's defenders are nowhere to be seen.

23 bxc3 ltJa4 24 i..xd4 The threat was . . . Vb6+. The other attempt to

cover that square, 24 :h6, fails to 24 . . . ltJxc3+ 25 �cl ltJce2+.

24 ••. exd4 25 �c2 The line 25 cxd4 Vh6+ 26 Vh3+ Vxb3+ 27

axb3 ltJc3+ 28 �c2 ltJxdl 29 �xdl l:tfl + leaves

w

White a piece down, as does 25 Vd3 Vh6+ 26 l:.b5 Vxb5+ 27 Vxb5 ltJxc3+. The king's attempt to flee fares no better.

25 ... Vf7 More incisive was 25 . . . Vb6 26 l:tbl :f2+ 27

Vxf2 d3+, but in such a position enough is as good as a feast.

26 l:txd4 26 J:r.d5 ltJxc3 allows White to make a few

moves more than in the game. 26 .•. Vxa2+ 27 �d3 'ii'b1 + 28 �d2 'ifb2+

0-1

Game 6 [085]

Alexander Lastin - Valeri Yandemirov Voronezh 200 7

1 d4 lLlf6 2 c4 g6 3 ltJc3 d5 4 cxd5 ltJxd5 5 i..d2 (D)

B

This seemingly passive move pursues the concrete goal of preparing to recapture on c3 with the bishop in order to neutralize its powerful adversary on g7. Meanwhile White intends e4, setting up the broad centre typical of the

Exchange Variation but without the structural liabilities. Should Black decline the offer to exchange and instead retreat the knight, White reaches standard positions with a tempo lost on the bishop move; then the question is how this tempo affects the assessment of the resulting situation. In this respect, the line is similar to the one seen in the previous game, where although White didn't lose a tempo with this bishop, the move f3 was of questionable usefulness, and indeed a lost tempo if White played a subsequent f4.

The 5 i..d2 line is old, but it has gained some popularity in the last few years.

5 ... i..g7 We should note that 5 ... i..f5? fails to prevent

the advance: 6 e4 ltJxc3 7 i..xc3 i..xe4?? 8 d5 and 9 Va4+.

The voluntary retreat 5 . . . ltJb6 is the main alternative, by which Black tries to prevent


White from reaching the most favourable setup - the advance e4 coupled with the fairly secure protection of the d4-pawn with .i.e3. The common continuation then is 6 ..ig5 and Black can gear up for . . . c5 according to his taste; he can: • first drive the bishop away to h4 or even g3; • push ... c5 before castling, so that the e7-

pawn is protected, and after dxc5 go into the endgame - exchange on c3, then the knight jumps from b6 to a4;

• play . . . c5 after castling and retreat the knight to d7, keeping the queens on the board. Of interest is another set-up where White

aims to forestall . . . c5: 6 tbf3 .i.g7 7 .l:.cl tbc6 (7 . . . 0-0 allows White to develop the bishop to an active post: 8 .i.g5 h6 9 .i.h4 g5 10 .i.g3 c5 1 1 d5 e6 1 2 d6, Morozevich-Shirov, Pamplona 2006) 8 e3 (in reply to 8 .i.f4 Sashikiran has successfully tested 8 .. . tbxd4 9 tbxd4 e5) 8 . . . e5 9 .i.b5 0-0 10 .i.xc6 exd4 1 1 tbxd4 bxc6 and the open lines and diagonals are quite adequate compensation for the pawn weaknesses, DreevKumosov, European Ch, Dresden 2007.

6 e4 tbxc3 The most common move is 6 . . . tbb6. After

the natural 7 .i.e3 0-0 (D) (delaying castling by 7 . . . tbc6 gives White the option of 8 d5 tbe5 9 f4 tbg4 10 ..ib5+ .i.d7 1 1 'ii'xg4 .i.xc3+ 1 2 bxc3 .i.xb5 1 3 .i.d4 or transposition to line 'b2' after 8 .i.b5 0-0) White has several ways to treat the resulting position:

w

a) 8 .i.e2 bears close comparison with line 'b' in which White first provokes . . . a6. After 8 ... tbc6 White has:

a l ) 9 tbf3 .i.g4 10 d5 (having allowed the enemy bishop on g4, White doesn't have sufficient resources to maintain the pawn on d4 and

must advance it) 1 0 . . . .i.xf3 1 1 gxf3 tba5 12 .i.d4 'ii'd6 is viable for Black.

a2) 9 d5 tbe5 10 .i.d4 is well met with 10 . . . c5 1 1 .i.xc5 tbec4 12 'ii'b3 tbxb2. Naturally, with the pawn on a6 the b6-knight would hang, so this solution isn't available to Black in line 'b' .

b) The somewhat abstract bishop sortie 8 .i.b5 (D) aims to provoke . . . a6, and then return to e2, claiming the inclusion to favour White, as it weakens the b6-square, and we have already seen in line 'a' how this can make a vital difference if Black opts for the natural moves. However, it is not so easy to find a useful reply that doesn't to some extent justify the bishop's position on b5:

B

bl) 8 ... .i.d7 interferes with the pressure on the pawn on d4, and White again retreats the bishop to e2.

b2) 8 . . . tbc6 allows White to defend d4 with 9 tbge2 without having to fear the pin that we saw in line 'al ' ; e.g., 9 . . . .i.d7 10 0-0 a6 1 1 .i.xc6 .i.xc6 12 d5 .i.d7, Korotylev-Timofeev, Kazan 2005. Instead, 9 .i.xc6 spoils Black's pawn-structure, but at the cost of opening for him the b-file and an important diagonal for the light-squared bishop. _

b3) 8 . . . .i.e6 attempts to create counterplay against c4.

b4) 8 . . . f5 is the principal method of undermining the white centre and should be compared with 'b5 ' .

b5) When Black does opt for 8 . . . a6 9 .i.e2, another typical plan of counterplay is preferred today: 9 . . . f5 with the possible continuation 10 exf5 .i.xf5 1 1 'ii'b3+ �h8 1 2 tbf3 tbc6 13 d5 tbe5 14 tbxe5 ..ixe5.

c) 8 h3 (D).


B

As mentioned in the notes to the previous game, such positions with an extra tempo for White are in his favour; naturally, the additional move grants Black more options in this case. Here, too, White can successfully withstand the piece pressure on the d4-pawn, so Black must undermine the centre with a pawnthrust. Both methods can be encountered: 8 . . . e5 9 ltJf3 exd4 1 0 .i.xd4, striving for an edge, and 8 . . . f5 9 exf5 ..txf5 1 0 ltJf3 ltJc6 II 'iib3+ �h8 12 d5 ltJe5 13 ltJxe5 .i.xe5 - we mentioned a similar position earlier. Although the above lines are not the only ones, they are the most typical.

d) 8 f4 transposes to the note to White's 8th move in Game 5.

7 .i.xc3 (D)

B

It was for the sake of this recapture that White initiated the bishop manoeuvre. Now the bishop effectively neutralizes its opposite number on g7, and in the likely case of the exchange of these bishops, Black is saddled with darksquared holes around his king. While these weaknesses may not be terminal, considering the forces remaining on the board, they require Black's constant attention.

7 ... c5 Black may postpone the counterblow in the

centre until his development is complete; e.g., 7 ... 0-0 8 .i.c4 ltJd7 9 ltJe2 c5 10 d5 ltJe5 1 1 .i.b3 e6 1 2 0-0 c4 1 3 .i.xc4 ltJxc4 14 ..txg7 �xg7 15 'ifd4+ 'iff6 16 'ifxc4 exd5 17 exd5 'ifxb2, Grishchuk-Krasenkow, European Clubs Cup, Saint Vincent 2005, although White's position is the more pleasant. The immediate blow in the centre also leads to positions that are more pleasant for White.

8 dS .i.xc3+ 9 bxc3 (D)

B

A new pawn configuration has arisen where White has a pawn more in the centre, which in the middlegame is of more consequence than Black's queenside majority. The resulting pawnchain leaves the c8-bishop with almost no place to go. Black must strive to restrict the freedom of the enemy soldiers and determine the final formation; this can be achieved by preparing . . . e6. Then the attempt to reinforce the d-pawn by c4 renders White's own bishop too passive, and . . . e5 with the idea of blockade becomes an attractive option. However, if White ends up recapturing with his bishop on d5, he will find the extra support provided by the c4-pawn quite handy. When castling kingside, Black must remember that his dark squares are weak and that . . . e6 aggravates this weakness further still, so he must constantly watch out for the h4-h5 attack.

9 .•. 0-0 Also seen is 9 . . . 'ifa5, trying to organize some

action on the queenside, but then the queen is farther away from the king, and Black is left with fewer defensive resources on the other side of the board.

lO l:i.bl


This move is inspired by the analogous manoeuvre from the l:tb1 Exchange Variation. The rook evades a possible attack and sets its sights on the b7-pawn. After 10 ltJf3 Black achieved equality in Karpov-Timofeev, Russian Team Ch, Sochi 2007: 10 . . . e6 1 1 .i.c4 b5 1 2 .txb5 'iia5 1 3 .i.c4 �xc3+ 14 ltJd2 .i.a6 1 5 :c1 'ife5.

10 •.• e6 Black decides to clarify the situation at once;

he is OK with the exchange of the b7 -pawn for the one on a2 that would now occur in the case of 1 1 dxe6 .i.xe6 12 �xd8 l:txd8 1 3 l:txb7 .txa2. Speaking formally, the change in the pawn-structure favours White: he has now one pawn-island fewer whereas Black has one more, although whether you consider the a7-pawn a passed pawn or a weakness depends on your reserve of optimism. In any case, Black is better developed (the a8-rook is already in the ideal place to support the advance of its ward). White prefers to reply in a classical fashion.

ll .tc4 �f6 The sortie 1 1 . . . 'iig5 is also possible. In reply

to the probable 1 2 �fl . Black can drop the queen back to f6, but it isn't wholly clear who benefits more from Black's little zwischenzug. The position of the king on f1 has its advantages - in some lines the pawns on c3 and e4 can be sacrificed without fearing a check, and the loss of castling is of little consequence if the rook can be activated along the h-file after h4-h5.

12 ltJe2 ltJd7 (D)

w

The weakness on e6 that appears after this move promises little joy, but this is the lesser evil. The alternatives are taking on d5, when the bishop takes the pawn's place, or closing the position with . . . e5, which in the given configuration is to White's advantage, as he still has the

diagonal for the bishop and the c4-square for the knight.

13 dxe6 fxe6 14 0-0 ltJb6 15 .i.b3 �e5 Why not 1 5 . . . c4 at once? Because of 1 6 .tc2

e5 17 a4, when it is still hard to complete development. White threatens a5, while the bishop must guard b7, and the c4-pawn is indirectly under attack as well, because of the possible 'iid5+. Denying the white queen that option by means of . . . :f8-d8 is undesirable, as that square is better suited for the a8-rook. That's why Black begins by transferring his queen with gain of tempo to c7, where it protects both pawns. However, the white queen also finds an active post.

16 'iid3 'ifc7 17 'ife3 c4 18 .tc2 e5 (D)

w

The initial impression is that Black is doing very well - the bishop has an open diagonal to operate on, while its opposite number has almost no moves; also, the e5-pawn restricts the enemy knight. However, this is all outweighed by the weak defence of the royal residence, and the threats that arise after f4 give White a certain advantage.

19 �g5 There were other interesting, and probably

stronger, continuations - 19 a4, 19 I;lb5 or 1 9 f4 at once. The text-move also looks attractive -the queen is poised for assault, controlling a complex of weak squares and removing itself from its position on the same diagonal as the king, in order to exclude the possibility of the exchange after . . . 'ifc5 in certain variations. White threatens the h4-h5 raid, while also setting his sights on the e5-pawn. Nevertheless, White still needs to bring up reinforcements if his attack is to carry much weight.

19 .•• ltJd7


It would be better to let the bishop out frrst, except that on e6 it would come under attack after the planned f4. However, the simple 19 . . . .td7 looks good - the mobilization nears completion, and the squares a4 and b5 are placed under control. The lines 20 f4 l:.ae8 2 1 fxe5 (21 f5 !?) 2 l . . . .Uxfl + 22 l:.xfl 'i¥xe5 and 20 h4 .Uf7 2 1 h5 .:te8 22 hxg6 hxg6 23 'i¥xg6+ l:.g7 24 'ii'h6 'ii'c5, with substantial compensation for the pawn, both look quite acceptable for Black. Instead Black prefers to reinforce the point e5 and develop the bishop to b7; if he manages both, he will have nothing to complain about.

20 .Ubdl (D)

B

20 .•. b6 20 . . . lt:Jf6 may be good, to meet 21 f4 with

2 1 . . .'ii'b6+ and 22 . . . lt:Jg4; then 2 l l:f.d2 secures White control of the d-file for a while, and he could attempt to take advantage of the opportunity to invade on d6.

21 f4 It seems that White is in a hurry to force the

pace; it was a good idea to loosen Black's defensive line with an advance of the h-pawn first.

21 ••• 'ii'c5+ 22 �hl exf4 23 l:f.d5 'i¥e3 24 lt:Jxf4

Nothing scary is to be seen after 24 . . . 1Wxc3 25 .:tf5 .tb7 26 lt:Jxg6 l:.xf5 and now 27 lt:Je7++ �h8 28 lt:Jxf5 l:tf8 or 27 'i¥xf5 hxg6 28 'i¥xd7 'ii'xc2 29 'ii'f7+ �h8 30 'ii'xb7 l:f.d8. However, the pin on the knight looks so frrm that Black decided to go for more and overlooked something.

24 •.• lt:Jc5? 25 .l:ld8 �b7 26 lt:Jh3! (D)

B

And here is the surprise, ending the game at once: Black's losses are too heavy. The following moves were not really necessary.

26 ... l:.axd8 27 :xr8+ :xes 28 'ii'xe3 1:.0 + 29 lt:Jgl �c8 30 h3 l:f.al 31 �h2 l:.xa2 32 �dl aS 33 'ii'd4 �e6 34 lt:Jf3 1-0

Game 7 [085]

P ierre Yves Toulzac - Valentin Bogdanov Malakoff 2007

1 d4 lt:Jf6 2 c4 g6 3 lt:Jc3 d5 4 lt:Jf3 Our topic in this game is the idea of White ex

changing on d5 followed by the odd-looking B move lt:Ja4. In the game, this occurs after the preliminary 4 lbf3 �g7, but naturally we should also examine the immediate 4 cxd5 lbxd5 5 lt:Ja4 (D).

When seen for the first time, this continuation produces much the same impression as 1 a4 or 1 h4; that is, at best it looks like a provocation. One can imagine what a fine target of


sarcasm this move would have made in the days of good old Dr Tarrasch. The basic opening laws are being violated - the knight spends another tempo . . . on withdrawing from the centre. However, in the past hundred years the assessment criteria have become more liberal and concrete. How can the exchange on c3 be avoided? Withdrawing from c3 is one possible answer. But withdrawing where? Not to e4 -White's plans call for that square to be reserved for a pawn. That leaves us with a4 - and at least there it does something useful by controlling the c5-square.

Of course, this continuation does not purport to refute the Griinfeld Defence; on the other hand, Black's attempts at counter-refutation, which this knight leap initially provoked, met with little success, and today this variation has the reputation of being quite playable.

Black's sharpest response is 5 . . . e5. Now 6 e4 tt::lf6 promises little, so White normally continues 6 dxe5 . Here two continuations have been tested:

a) 6 . . . ..tb4+ 7 ..td2 tt::le3 (D).

w

This tactical trick is the idea behind Black's counterplay. Now in reply to 8 1i'c 1 the simple 8 . . . ..txd2+ 9 1i'xd2 1i'xd2+ 10 ..ti>xd2 tt::lc4+ is good enough, so 8 fxe3 ..txd2+ 9 1i'xd2 (9 'Ofi>f2 tt::lc6 10 tt::lf3 ..tb4 offers White no advantage) 9 . . . 1i'h4+ 10 g3 1i'xa4 is the main line. In the stem game for the tt::la4 idea, the system's author Nadanian chose 1 1 1i'd4 1i'a5+ 1 2 b4, but later it was discovered that after 1 2 . . . 1i'a3 Black stands well. Today the plan of 1 1 ..tg2 followed by 1i'c3, 0-0 and play on the weak dark squares and half-open f-file is preferred. The tripled pawns on the e-file can hardly constitute a material advantage, but White's initiative persists,

and this demands particularly accurate play from his opponent.

b) With 6 . . . tt::lc6 Black puts off winning back the pawn and aims at rapid development, setting his sights on the insufficiently protected c2-square. One possible course of events is 7 tt::lf3 tt::ldb4 8 ..tg5 'i!Vxd 1 + 9 'Oti>xd 1 ..te6 10 tt::lc3 ..tg7 1 1 a3 tt::ld5 12 e4 tt::lxc3+ 13 bxc3 tt::lxe5 with a roughly even game. If White forestalls the leap to b4 and returns the pawn at once, at the same time depriving Black of the right to castle, his achievements are still rather modest: 7 a3 tt::lxe5 8 e4 tt::lb6 9 1i'xd8+ 'Oti>xd8 10 ..tg5+ ..te7 1 1 0-0-0+ 'Oti>e8 12 ..tf4 ..td6, Bu XiangzhiSashikiran, Dos Hermanas (Internet Knockout) 2005.

Let us suppose that Black seeks a creative response. Can he also spend a tempo on moving the knight twice, in order to prevent e4? Indeed he can, and the reply 5 . . . tt::lf6 is quite solid. Then building the ideal centre cannot be secured by natural means, so White usually transposes into variations with the fianchetto of the light-squared bishop or posts the dark-squared one on f4, acquiescing to a less ambitious formation in the centre for the time being: 6 tt::lf3 ..tg7 7 ..tf4 tt::lbd7 8 h3 0-0 9 e3 c5, Dao Thien Hai-Wang Yue, Doha (rapid) 2006.

Finally, 5 . . . ..tg7 usually transposes into the position in the game; there have been few attempts to develop the knight to e2.

4 .•. ..tg7 5 cxd5 tt::lxd5 6 tt::la4 (D)

B

6 ••• 0-0 White's move-order has avoided the lines

with . . . e5, at the cost of giving Black 6 . . . ..tf5 as an extra option. With the knight still on g l , the natural reaction would be f3, but here White must take steps to deal with the knight leap to


b4. Earlier, the continuation 7 lt:lc5 b6 8 e4 bxc5 9 exf5 cxd4 was the subject of practical tests; today the reply is 7 lt:lh4 with an amusing knight symmetry.

7 e4 lt:lb6 White will now generally place his bishops

on e2 and e3, but there are some subtleties relating to the order in which he makes these moves.

8 i.e2 By developing this bishop first, White avoids

the immediate . . . c5 that is possible, though hardly sensible, in case of 8 i.e3, i.e., 8 . . . lt:lxa4 9 'ifxa4 c5. This line doesn't enjoy a particularly great reputation, since after 10 l:td I it is not easy for Black to neutralize the activity of White's pieces. 8 i.e3 also gives White an additional possibility after 8 . . . i.g4, namely 9 lt:lc5 (avoiding yet another version of the exchange on a4 with a subsequent . . . c5; 9 i.e2 obviously transposes to the main game), although this has its pluses from Black's point of view as well: 9 . . . lt:lc6 10 lt:lxb7 'ifc8 I I i.a6 lt:lb4. With the continuation in the game, the order in which the bishops are developed has no importance.

8 ... i.g4 9 i.e3 lt:lc6 Here the already mentioned idea of 9 . . . lt:lxa4

1 0 'ifxa4 c5 1 I .l:td l cxd4 is sounder, as simplification offers Black better reasons to hope for equality after both 12 i.xd4 i.xd4 1 3 .l:txd4 'ii'b6 and I2 lt:lxd4 i.xe2. The text-move leads to another typical Griinfeld set-up.

10 d5 (D)

B

10 ... i.xf3 The attempt to exchange the b-pawns makes

little sense: I O . . . lt:lxa4?! I I 'ifxa4 i.xb2 I 2 .l:tbl i.c3+ 1 3 'it>fl lt:lb8 ( 1 3 . . . i.xf3?! I 4 i.xf3 lt:\e5? loses a piece to I5 'ifa3) I4 'ii'b3 and I 5 'ifxb7. However, the simplifying I O . . . lt:\e5 I I

lt:lxb6 axb6 I 2 lt:\xe5 i.xe2 1 3 'ifxe2 i.xe5 is solid enough.

1 1 gxf3 lt:\e5 Although such play leads to complex posi

tions, it is a risky strategy. White has a strong centre and two bishops, which promises an advantage in many endgames. The price is a slight defect in the pawn-structure that poses the problem of choosing a place of residence for the white king. Still, the potentially weak h-pawn can be thrown into the assault by h4-h5, and the king's freedom to choose his residence provides considerable flexibility in meeting Black's plans. Black's top priority is sapping the d5-pawn, which is not easy to arrange.

Although a retreat at the cost of development is unpleasant, I I . . .lt:lb8 deserved consideration, with the possible sequel i2 'ii'b3 lt:lxa4 1 3 'ifxa4 c6 I 4 0-0-0 cxd5 I5 .l:txd5 'ii'c7+ I6 'it>bi lt:lc6.

12 .l:tcl (D)

B

12 ... c6? A thoughtless move, as Black now simply

loses a pawn. Preferable was I 2 . . . lt:led7 1 3 lt:lc3 c6 14 dxc6 bxc6 15 f4 'ii'b8 16 h4 l:td8, Egiazarian-Neverov, Minsk 1998, even though here too, White has the better chances.

13 f4 lt:led7 14 dxc6 bxc6 15 .l:txc6 .l:tc8 Maintaining the material parity is even worse:

after 15 . . . lt:lxa4 I6 'ifxa4 i.xb2 17 0-0 White's dominance will quickly bring him new acquisitions. Black, a pawn down, shifts into difficult defence and for a start exchanges off the active white rook.

16 .l:txc8 'ifxc8 17 0-0 'ifc6 18 lt:lxb6 lt:lxb6 (D)

Time to sum l!P the results: they are sad for Black - there is no compensation for the pawn.


w

Were his forces active, the somewhat open position of the enemy king might give him a glimmer of a hope, but even that is missing. On his last move Black had to choose which of the remaining minor pieces should be allotted more scope. The capture on b6 by the pawn would keep some options for the knight open, but then White would block the bishop with e5. Now this advance is inexpedient, as it would allow the knight to d5, so the bishop is not 'bad' for the time being. However, after White's b3 it will be shut out of the fight.

19 'ili'b1 e5 In order to thin out the pawn-formation

around the enemy king and provide a future for his own bishop. If 19 . . . ltJa4, then 20 e5.

20 l:tcl 'ili'f6 Now Black is forced to recapture on e5 with

the queen, and the transfer of the bishop to the same diagonal as the h2-pawn has to be postponed. However, Black didn't like 20 . . . 'i!i'e6 because of 2 1 f5 and both black minor pieces would have been inactive for some time.

21 fxe5 'i!i'xe5 22 b3 'i!i'e7 23 �f3 White does not hurry. He had a good choice

of active continuations; for example, 23 f4 or 23 a4. At the moment he has the more active pieces in addition to the extra pawn; should Black manage to neutralize this factor, his hopes for salvation would increase.

23 ... �e5 24 �g2 �b8 (D) Black tries to worry White with ideas of the

queen invading on h2, even though this isn't much of a threat.

25 �c5 �d6 26 �xd6 This exchange serves no purpose. The essen

tial condition of the advantage of two bishops consists first of all of the fact that there are two of them. Now the dark squares are weakened.

w

Still, Black's achievement is of a rather symbolic nature, as White's position remains won.

26 ••. 'i!i'xd6 27 :c2 White embarks on a slow regrouping plan;

he should have taken care that the enemy knight remained shut out of the game instead. 27 l:tdl controlling the main transit point d7 suggested itself.

27 ••. ltJd7 28 'i!i'cl ltJe5 29 l:td2 'i!i'b8 30 'i!i'd1 l:tc8

The regrouping has ended in the complete and absolute control of the d-file, but the value of control over an open line is the highest when that open line is the only one. If your opponent has an open line of his own, then you should produce the proof that yours is more important.

31 f4 (D)

B

The pawns relocate to dark squares, to widen the bishop's scope of activity. This is a sensible decision with one drawback - the area around the king is weakened still more.

31...ltJc6 32 e5 ttJe7 33 �h3? 'i!i'b6+ 34 'it>g2 l:tc3 35 l:td3 'i!i'c6+

White's last few moves were not best (the bishop on h3 is more of a hindrance than a


workhorse, and it was not necessary to allow, even briefly, a pin on the rook), and Black has acquired a serious initiative.

36 'it>g3? 36 .l:r.f3 was mandatory, even though the con

tinuation 36 . . . 1!i'e4 37 'iil>f2 l::f.c2+ 38 'it>gl looks dangerous. White should hold the game, as he will play for a perpetual check on the squares d8, f6 and h4, in case Black attempts to activate the knight.

36 ••• l::f.cl? Black misses an opportunity to demonstrate a

classic example of a double pin by 36 . . . 1!i'd5 (D).

w

The only reply, 37 �fl . is met by 37 . . . tt'lf5+ 38 'it>h3 'ii'e6 39 'ili'f3 (giving up the exchange with 39 'it>g2 tt'le3+ offers as few prospects of salvation) 39 . . . tt'le3+ with a multitude of threats that White would be hard pressed to meet; for example, 40 'it>g3 tt'lxfl+ 4 1 'ili'xfl 'ili'a6; 40 'it>h4 'ii'e7+ 41 'it>g3 (41 'it>h3 .l:r.xd3 42 �xd3 'i!i'd7+) 4 l . . .tt'lf5+ 42 'iil>f2 'ii'c5+ 43 'it>e1 .l:r.c l + 44 .l:r.d 1 .l:r.c2; or 40 f5 'iii' xeS - only this line gives a glimmer of hope.

37 'ii'f3 .l:r.gl+ 38 �g2? tt'lf5+ 39 'it>h3 (D)

B

39 ••• 'ii'c8?

Having spent the last few seconds before the time-control and found nothing decisive, Black forgets about the presence of the mighty bishop on the long diagonal and makes a horrible move. 39 . . . 'ii'c2?, hoping for 40 e6? .l:r.xg2 4 1 .:.d8+ 'it>g7 42 'ii'xg2 'ili'c3+, also fails to win, as White interpolates 40 .l:r.c3, and if the queen goes to b2, it loses the chance to check along the third rank. Nevertheless, 39 . . . 'ii'e6 40 'ii'd5 tt'ld4+ 41 'ii'xe6 l2lxe6 42 .l:r.f3 g5 was good, but Black failed to find this at the board and decided to create a threat from another square, overlooking White's obvious reply.

40 'ii'a8 Black is back in the hole, except it is even

deeper now. The unfortunate position of the white king is his only chance.

40 ••• 'ili'xa8 41 �xa8 :n 42 .l:r.d7 .l:r.xf4 43 .:.xa7 g5?! (D)

The wrong pawn. The rook endgame offered real chances of salvation: 43 . . . h5 44 .l:r.a4 .l:r.f2 45 �e4 g5 46 �xf5 .l:r.xf5 47 .l:r.a5 .l:r.f3+ 48 'it>g2 .l:r.c3 49 e6 fxe6 50 .l:r.xg5+ 'it>f7.

w

44 .:.a4 .:.f2 45 l::tg4 h6 46 .l:r.g2 The rook comes to the rescue in time, picking

up a pawn along the way. Now the pair of passed pawns and the strong bishop make White's technical task quite feasible.

46 ••• :n 47 �e4 tt'ld4 48 .l:r.d2 tt'le6 49 �d5 For the sake of keeping the e5-pawn, White

agrees to the exchange of the minor pieces -and wrongly so. With the bishop present the passed pawns would run at a much faster pace.

49 ••• l2lf4+ 50 'it>g4 tt'lxd5 51 .l:r.xd5 .:.r2 (D) And here is the rook endgame, along with its

inherent drawish traditions. White still has a connected pair of candidates for joint promotion, who will advance a step at a time while


w

Black is busy acquiring the same. In this sort of race, the decisive factor is how far the pawns are advanced. The only other important variable is the position of the king facing the enemy foot-soldiers - either he will manage to slow down their advance or, which is more likely, it will prove a welcome target for their fire and present Black with the much-needed tempi.

52 a4 l:txh2 53 a5 l:tg2+ 54 �f3 l:ta2 55 b4 �g7 56 l:td6?!

56 l:i.d7 or 56 �e4 offers better chances of victory.

56 ... h5 57 a6 h4 58 �g4 l:tg2+ 59 �h3 (D) Insufficient was 59 �f5 h3 60 l:td I l:ta2 61

b5 lla5 with mutual extermination.

B

59 •.. l:tg3+ 59 . . . l:tb2 blocks the king's way forward be

cause of the threat of taking with check on b4 and provokes an advance of the a-pawn, after which that pawn would for some time be left without the support of its neighbour and thus tie down the rook to its defence: 60 l:td8 l:tb3+ 6 1 �g2 l:tb2+ 62 �g1 l:tb1 + 6 3 �f2 l:tb2+ 64 �e3

l:ta2 65 b5 h3 or 60 a7 l:tb3+ 61 �g2 %:tb2+ 62 �f3 l:!.a2 63 l:td7 h3. This is probably a more reliable way to hold the draw.

60 �h2 l:ta3 61 b5 l:!.a5 62 l:td8 The same reply would follow after 62 l:ld7,

with this independent line: 62 . . . g4 63 e6 l:.a2+ 64 �g1 l:tal + 65 �g2 h3+ 66 �g3 %:tg1 + 67 �h2 l:!.g2+ 68 �h1 lla2 69 llxf7+ �g6 70 l:tf8 �g5.

62 • • . g4 63 b6 After 63 l:td4 l:txb5 64 l:lxg4+ �h6 65 l:!.xh4+

�g5 66 l:ta4 l:tb8 67 a7 l:ta8 68 �g3 �f5 69 l:ta5 f6 Black would also manage to save his scalp.

63 ... .l:.a2+ 64 'it>gl l:tal + 65 �f2 g3+ 66 �3 l:tn + 67 �g2 l:.f2+ 68 �gt hJ ( D J

w

A picturesque position, illustrating the superiority of the pawns charging the king. Although it is White's tum to move, it is he who would lose were the enemy passed pair shifted one file to the left. Here, however, the black rook doesn't have enough room, and the players had to light the pipe of peace.

69 l:tdl l:tg2+ 70 �hl l:th2+ 71 �gl lfl.lfl

The variations examined in this chapter have the laudable aim of augmenting the effect of the e4 advance, but none are free of obvious drawbacks. Therefore it is only logical that they should not render the opening's state critical. Nevertheless, a fresh, quality idea can secure short-term success and cause the popularity of a particular variation to surge, which is what happens today in the systems from Games 5 and 6.

3 Classical Exchange Variation : White Plays 'iJe2

We shall now move on to the Exchange Variation, in which White takes advantage of the opportunity offered to him to seize the centre the most natural way: 1 d4 lbf6 2 c4 g6 3 lbc3 d5 4 cxd5 lbxd5 5 e4 lbxc3 6 bxc3 (D).

B

Having allowed White to place the pawns in such a pretty manner, Black has acquired a certain initiative. And indeed he needs this initiative if he is not to be over-run. Black's counterattack against White's centre pawns will limit the choice of squares for the development of the white pieces, whose main concern must be the preservation of the central formation.

The white knight has two reasonable deployments. This chapter discusses lines where the knight is developed to e2, where it is safe from the pin by the bishop on g4, as . . . .tg4 can be met by f3. This method of play dominated the theory and practice of the Exchange Grtinfeld for a long time, but for a couple of decades, starting in the late 1 970s, it went out of fashion, eclipsed by the lLlf3 lines that we examine in the next chapter. In the last decade, new resources and ideas have been found in the lbe2 systems, while the theory of the lLlf3 lines became increasingly worked out. Today there is an approximate parity between the two set-ups.

After the further moves 6 . . . .tg7 7 .tc4 c5 8 lLle2 lLlc6 9 .te3 0-0 I 0 0-0 (D) the critical position arises.

B

CLASSICAL EXCHANGE VARIATION: WHITE PLAYS ti:Je2 39

Here Black has to pick one of three main lines.

In Game 8 (Nybiick-Krasenkow) Black chooses to fianchetto his queen's bishop as well, with subsequent play against the white pawns in the centre and operations along the newly-opened cfile.

Game 9 (Aronian-Shirov) is devoted to the traditional main line where Black uses piece-play in the most direct manner to try to break up White's centre or else to create counterplay by using the c4-square. The move . . . .i.g4 figures prominently in this plan, despite the fact that it can be met by f3, because this move weakens an important diagonal and White's second rank. The spotlight of the subsequent struggle is focused on a well-known variation featuring an exchange sacrifice. Although more than half a century old, this line has recently obtained a new lease of life as it is particularly susceptible to computer analysis, which has uncovered a vast array of new ideas for both sides.

The chapter closes with Game 10 (Kobaliya-Areshchenko), which examines complex multifaceted schemes of creating pressure on the white centre. These used to be based on the set-up with 10 . . . 'ilfc7 and .. J:td8, but today they are generally initiated with the development of the bishop to d7.

Game 8 [088]

Tomi Nyback - Michal Krasenkow European Ch, Antalya 2004

1 d4 lLlf6 2 c4 g6 3 lLlc3 d5 4 cxd5 lLlxdS 5 e4 lLlxc3 6 bxc3 .i.g7

The order in which . . . c5 and . . . ..tg7 are played has minimal importance and in most games leads to a transposition. The main exceptions are those cases where White goes with the alternative development of the bishop to b5 or when Black plans to carry out . . . c5 much later, after a partial mobilization.

7 .i.c4 7 .i.a3 has long ceased to be feared - the

bishop is not posted particularly well there, and it cannot in any case prevent Black from preparing . . . c5.

Let's examine 7 .i.b5+ (D) in combination with the subsequent development of the knight to e2.

The point of the check is to post the bishop on a4 instead of c4. There is some benefit to this - the bishop avoids a possible attack along the c-file, but there are drawbacks as well - White's control over the c4-square is decreased, and Black gets additional options connected with the appearance, albeit brief, of a pawn on c6. One such option is the counterattack in the

B

centre with . . . e5, since the advance o{ the d4-pawn would now be a dubious idea; another aims at the seizure of space on the queenside by means of . . . b5, . . . aS, etc. In the classic set-up the counterplay is based on striving to take advantage of the c4-square: 7 . . . c6 8 .i.a4 0-0 9 lLle2 c5 I 0 0-0 lLlc6 I I .i.e3 lLla5. There is an alternative version as well with 7 ... .i.d7, when White usually chooses 8 .i.e2, as the exchange of the bishops on the one hand reinforces the white pawn on d4, as the black bishop is of more use in the attack of this pawn than the white one in its defence, but on the other hand,


the simplifications are good for Black. Having lured the bishop out to d7, White has restricted the choice of the variations available to Black B in the system with the knight on f3, but opened to him an additional original option: 8 . . . c5 9 tZ:lf3 cxd4 10 cxd4 .i.c6 I I �d3 f5 12 exf5 �a5+ 1 3 .i.d2 �xf5 .

We now return to 7 .i.c4 (D):

B

7 ... c5 Attempts to postpone . . . c5 used to have some

popularity, but are very rarely seen nowadays: a) 7 . . . b6 gives White a respite from the work

in the centre and he can afford an immediate attack on the enemy king: 8 �f3 0-0 9 e5 .i.a6 10 .i.d5 (to take away the c6-square from the knight; 10 .i.b3 at once allows Black a counterattack with 10 . . . �d7 and I l . . .tZ:lc6) 10 . . . c6 1 1 .i.b3 �c7 12 tLle2 .i.b7 1 3 h4 c5 14 �g3 cxd4 1 5 cxd4 tZ:lc6 1 6 h5 tZ:lxd4 (Black hopes to obtain three pawns for the piece but there is an elegant refutation) 17 tZ:lxd4 �xe5+ 1 8 �xe5 .i.xe5 19 .i.e3 .l:.fd8 20 .l:.d 1 .i.xg2 21 .l:.g 1 .i.d5 22 hxg6 hxg6 23 tZ:lc6! .i.xc6 24 .l:.xg6+ and 25 .l:.xc6, winning, Truskavetsky-Shishkin, Ukrainian Ch, Simferopol 2003.

b) 7 ... tZ:lc6 leads to analogous positions - 8 tLle2 b6 9 h4 with the standard initiative.

c) 7 . . . 0-0 leads to a transposition of moves in most games; however, now White, too, manages to castle, which deprives Black of two options, although ones that are not very popular. We shall consider them below.

8 tLle2 As mentioned above, the idea of the whole

set-up is based on this development of the knight, not allowing a pin.

8 ... tLlc6 9 .i.e3 (D) 9 ... 0-0

We mentioned above that the fact that White has not yet castled allows Black two (rare) ways to seek counterplay. These both feature the preliminary exchange 9 . . . cxd4 10 cxd4. Then:

a) 1 O . . . b5 . For those with peaceable inclinations the line 1 1 .i.xb5 �a5+ 12 tLlc3 �xc3+ 1 3 .i.d2 �xd4 14 i.xc6+ .i.d7 15 .i.xa8 �xal is a popular choice to register the result. However, White has a nice opportunity to fight for the advantage: 1 1 .i.d5 .i.d7 12 .l:.c 1 .l:.c8 13 0-0 0-0 14 .i.xc6 .l:.xc6 1 5 .l:.xc6 .i.xc6 1 6 d5 .i.d7 17 �d2 a5 1 8 .l:.c l , and by combining the attack on the queenside pawns with the transfer of the knight to c6 via d4 White maintains the better chances.

b) 10 . . . �a5+ (D) and now:

w

b1) The exchange of queens 1 1 �d2 �xd2+ 12 �xd2, leaving the king in centre, is OK with Black, as he has ways to disturb the harmonious arrangement of the enemy forces.

b2) After 1 1 'iir>fl White hopes that the exposed position of the black queen will compensate for the forfeit of the right to castle and the associated time loss, and that the h 1-rook can be considered developed once the h-pawn advances. However, the plan with the rapid use of

CLASSICAL EXCHANGE VARIATION: WHITE PLAYS liJe2 41

the c4-point with . . . 'i!ka3, . . . lt:'Ja5 and . . . .i.e6 doesn't look bad for Black.

b3) 1 1 .i.d2 looks more natural. Then: b3 1 ) Black can take comfort in the transfer of

the bishop from e3 to c3 (where it is active, in a way, yet less secure) and retreat with l l . . .'i!kd8. In the line 12 d5 tt:'le5 1 3 .i.c3 we should note the game Topalov-Mamedyarov, European Team Ch, Khersonisos 2007: 1 3 . . . .id7 (the threat of f4 is usually parried with castling; here Black solves this problem with the counterattack along the c-file) 1 4 .i.b3 .l:.c8 1 5 l:.cl 'i!kb6, and here White found nothing better than 1 6 .i.a4 0-0 with simplifications, and Black didn't have any particular difficulties holding the position. The search for an advantage in the alternative continuation 1 2 .i.c3 0-0 1 3 0-0 was frustrated in another Mamedyarov game: 1 3 . . . .i.g4 14 f3 .i.d7 1 5 'i!kd2 a6 1 6 a4 .l:.c8 17 .l:.fc I b5 with a total emancipation, Van Wely-Mamedyarov, Hoogeveen 2007.

b32) 1 1 . . . 'i!kh5 is more ambitious. White again has a choice between reinforcing the centre and completing his development, viz. 1 2 0-0 0-0 1 3 .i.c3, and the immediate 12 d5 tt:'le5 1 3 .i.b5+ .i.d7 1 4 .i.xd7+ tt:'lxd7, at which point we should note the game Valsser - VachierLagrave, French Ch, Besan�on 2006: 1 5 .l:.b1 'ilk g4 1 6 tt:'lg3 'i!kxd 1 + 1 7 'it>xd 1 h5 1 8 f3 b6, when Black's accurate play was enough for equality.

10 0-0 (D)

B

10 . • . lt:'Ja5 With the system of development chosen by

Black, this move is the most accurate. The preliminary exchange 10 . . . cxd4 1 1 cxd4 tt:'la5 gives White the extra option of 1 2 l:tc 1 , although this continuation has few fans.

10 . . . b6 usually leads to a transposition of moves, but whereas earlier Black didn't have to worry about the capture on c5, as his expectations of regaining the pawn and leaving White with two weaknesses were fully justified, now this is a genuine sacrifice, and White can take the pawn, although Black does have a certain compensation after 1 1 dxc5 'i!kc7 1 2 tt:'ld4 tt:'le5 1 3 tt:'lb5 'i!kb8.

1 1 .i.d3 cxd4 The c-pawns could be kept on the board with

the immediate 1 1 . . . b6. In this version the acceptance of the sacrifice by 1 2 dxc5 bxc5 1 3 .i.xc5 'i!kc7 1 4 .i.d4 e5 1 5 .i.e3 tt:'lc4 typically ends with White, not Black, having to seek equality, which is why White prefers other continuations. One possible formation arises after 1 2 d5; Black then doesn't allow its neighbour on the c-file to bolster it and plays 12 . . . c4 himself, with the subsequent preparation of the undermining . . . e6. Another one received practical tests only recently: 12 .l:.c 1 e5 (Black tempts his opponent with the chance to obtain a protected passed pawn, but in the blocked position White's pieces would become passive) 1 3 dxc5 (since the pawn on e5 blocked the bishop's diagonal, White prefers to accept the sacrifice in this situation) 1 3 . . . .i.e6 1 4 c4 bxc5 1 5 .i.xc5 .i.h6 16 f4 .l:.e8 17 f5 with sharp play, Harikrishna-Sutovsky, Montreal 2007.

12 cxd4 (D)

B

On the board there is a standard formation, also found in other openings. White endeavours to use his superiority in the centre to launch a pawn advance and at the same time readies for the switch to an attack on the enemy king. Either pawn can initiate the advance forward. As a rule, exchanges don't favour the attacking


side, but that doesn't apply in the case of the defender on g7, who can be offered the deserved retirement either from h6 or from d4. The open c-file presupposes a partial or complete elimination of the rooks, since avoiding these exchanges would mean giving up this important file. For the time being Black pursues a policy of containment and decides on the order of . . . b6, . . . e6, . . . .i.b7 and . . . 'iVd7/d6. His trump the two-to-one majority on the queenside -can increase in value as the position is simplified.

12 ••• b6 13 'iVd2 If White begins with 1 3 l:tc I , then the pro

phylaxis against the advance of the d-pawn by means of I 3 . . . e6 is accepted as the most accurate continuation, while 1 3 . . . .i.b7?! harks back to an old game featuring spectacular tactics: 14 d5 e6 1 5 dxe6 fxe6 16 lt::lf4 'iVd6 1 7 'iVg4 l:tae8 1 8 .i.b5 l:te7 I 9 l:fdl 'aVeS 20 .i.d7 'ii'xe4 21 .i.xe6+ 'it>h8 22 l:td4!, Geller-Kapengut, USSR Ch, Leningrad 1 97 1 .

13 •.• .i.b7 14 l:tacl 'iVd7 (D) An attempt to save a tempo. The main contin

uation is 14 . . . e6, when White strives to weaken the walls of the enemy's royal residence with 1 5 .i.h6 or 1 5 h4. Here is a fresh example of youth defeating experience: 15 h4 'iVe7 1 6 h5 l:tfc8 17 e5 (a characteristic moment: by ceding d5 to Black, White gets a temporary respite from the pressure on his centre pawns, fixes the dark squares and concentrates on the kingside attack) 17 . . . l:txc l 1 8 l:txc I l:tc8 19 l:txc8+ .i.xc8 20 .i.g5 'iVc7 2 1 .i.f6 lt::lc6 22 'iV g5 h6 23 'aVe I , Carlsen-lvanchuk, Linares 2007. The queen's 'long' retreat adds to the assault on g6 the threat to win the pinned knight by .i.b5, d5 and lt::ld4.

w

15 'iVb4

White switches to the other flank and plans to harass the enemy queen.

15 ••• lt::lc6?! An unfortunate reply; 1 5 . . . e6 16 .i.b5 lt::lc6

I 7 'iVa4 l:tfc8 is better. 16 'iVa4 'iVd6 The attempt to transpose to the previous note

via 1 6 . . . e6 doesn't work, since White replies I 7 d5 at once.

17 .i.b5 lt::la5 This is forced, as I 7 . . . l:tfc8? loses to I S e5

'iVd7 1 9 d5 'iVxd5 20 lt::lf4 'iVd7 2 1 l:tfd I 'aVeS 22 lt::ld5, and the threats of multiple captures on c6 with the concluding fork on e7 or of piling up on the pinned knight with lt::lb4 cannot be met simultaneously.

18 e5 The advance by the other pawn, 1 8 d5 !?, ap

pears no less attractive. 18 ••• 'iVd8 19 l:tfd1?! (D) 1 9 .i.d7 ! , threatening e6, looks logical; for

example, 1 9 . . . .i.d5 20 lt::lf4 .i.xa2 2 1 .i.g4 and to save the bishop Black has to return the pawn with 2 l . . . .i.b3 22 'iVb4 b5, after which the difference in the position of the minor pieces speaks for itself.

B

19 ••• a6 20 .i.d3 b5 21 'iVb4 .i.d5 22 lt::lf4 .i.c4

The capture of the a2-pawn doesn't lose a piece, but it leaves the bishop stuck on a4: after 22 . . . .i.xa2?! 23 .i.e4 l:tb8 (or 23 . . . l:ta7) 24 l:tal .i.b3 and 25 . . . .i.a4, White's superiority in the centre should prove decisive.

23 .i.e4 l:tc8 24 'iVa3 l:te8 25 h4 A resource typical of such positions. 25 ••• e6 26 g3?! After 26 h5, Black must take into account

various versions of the sacrifice on g6, as well

CLASSICAL EXCHANGE VARIATION: WHITE PLAYS ti:'Je2 43

as the constant threat of h6. However, White stops halfway through and falls for another idea.

26 .•. ..if8 27 'ii'b2 'ii'e7 28 ..id2?! lbc6 (D)

w

29 d5? Spectacular yet bad. During the last few

moves Black has noticeably improved the centralization of his forces, so there should be no reason for such a sharp action.

29 ... exd5 30 .l:.xc4 dxe4? A gift in return. The bishop, capable of clear

ing the e-pawn's path, seemed the more dangerous combat unit; however, the refutation lies in 30 . . . dxc4 3 I lbd5 'ii'e6 32 lbf6+ 'ii'xf6 33 exf6 .l:.xe4 with real winning chances.

31 lbd5 'ii'e6 32 lbf6+ 'it>h8 33 llxe4 lled8 34 h5

Black thought that here too, the control of the centre files secures him a substantial advantage, but White has enough counterplay to maintain the balance.

34 ... ..ig7 35 .l:.del lbxeS 36 llxeS ..ixf6 (D) 37 ..igS

w

This move saves the game. The main threat is 38 ..ixf6+ 'ii'xf6 39 .l:.e8+ 'it>g7 40 h6+.

37 ... .l:.f8?! The simplest way to parry White's threat

was 37 . . . gxh5 38 ..ixf6+ 'ii'xf6 39 .l:.xb5 with an even rook endgame. After the text-move Black again has to fight for the draw.

38 .l:.xe6 ..ixb2 39 .l:.xa6 gxhS Allowing the pawn to h6 after 39 . . . .l:.c2 and

winding up with a passive king is also unpleasant.

40 .l:.bl ..id4 Worse is 40 . . . .l:.c2? 4 I .l:.xb2. 41 .l:.xbS f6 42 ..ih6? The wrong square. Even though the bishop

would have had no moves left, 42 ..ih4 was correct, tying Black down to the defence of the f6-pawn, which, coupled with the advance of the passed pawn, gives White considerable winning chances. Now Black activates his pieces and carefully works out a draw.

42 ... .l:.g8 43 'it>f1 .l:.c2 44 .l:.fS .l:.d8 45 a4 'it>g8 46 .l:.f3 .l:.dS 47 ..ie3 ..ixe3 48 .l:.xe3 .l:.fS 49 .l:.e2 .l:.xe2 50 'it>xe2 'it>g7 51 'it>e3 h4 52 gxh4 1h.-1h.

Game 9 [089]

Levon Aronian - Alexei Shirov Candidates match (game 5), Elista 2007

1 d4 lbf6 2 c4 g6 3 lbc3 dS 4 cxdS lbxdS 5 e4 lbxc3 6 bxc3 ..ig7 7 ..ic4 cS 8 lbe2 lbc6 9 ..ie3 0-0

In this game, we look at lines where Black embarks upon immediate counterplay involving the moves . . . ..ig4, . . . lba5 and, normally, . . . cxd4.

We should note that . . . ..ig4 before castling is pointless, as . . . lba5 is not an option with the c4-bishop indirectly defended by a possible check on a4.

10 0-0 .i.g4 (D) There is no reason to begin with I O . . . lba5, as

after I I ..id3 .i.g4 the d4-pawn is adequately


protected, and instead of the weakening 1 2 f3 White has at his disposal the useful 1 2 h3.

A few words about the timing of the ex- B change on d4. As a rule, the choice lies with Black, who gets to decide whether he is comfortable with the line 1 0 . . . cxd4 1 1 cxd4 i.g4 1 2 f3 lL'la5 1 3 .l:.c1 and which formation - with the c-pawns or without them - is preferable in case of the capture on f7.

w

Black keeps up the pressure on d4. ll f3 This move is virtually forced; the sacrifice of

the c3-pawn by 1 1 d5?! lLla5 12 i.d3 c4 is extremely rare.

ll ... lLla5 The logical sequel to the previous move - it is

desirable to secure the e6-square for the bishop with a view to exploitation of c4. How is White to react?

12 i.d3 This is the main line, but there are two im

portant alternatives to consider: a) 12 i.d5 forces the black bishop to retreat

to a less prestigious station by 1 2 . . . cxd4 1 3 cxd4 i.d7, but it expects to find gainful employment on the a6-fl diagonal, whereas the d5-bishop should be prepared to receive little hospitality from the enemy e-pawn. Nevertheless, lines like 14 .:tb1 a6 1 5 i.xb7 .l:.a7 1 6 i.d5 i.b5 typically result in the restoration of the material balance and in equality. Revisiting the question of the proper timing of the exchange on d4, we should note that with live infantry on the c-file, Black has an additional possibility in 12 . . . i.d7 (without exchanging on d4) 1 3 l:tb1 'i!i'c7; with the c-file open it would imprudent to place the queen on c7 or c8.

b) 12 i.xf7+ (D).

The interest in this variation arose after Karpov used it against Kasparov in the Seville match in 1 987, and in several subsequent games. The move is more than obvious, though it had been considered not to be dangerous - the extra pawn has very modest material value and doesn't promise much in the endgame, whereas the light squares - in particular, the important c4-square - are left without sufficient protection. However, Karpov was able to show that the gpawns do possess a certain functional value: one remains in the king's defensive service, while the other is enlisted in combat action, and the g7-bishop runs the risk of incarceration, hemmed in by the d4-e5 pawn-chain and the pawn on g5 . Such high-level use typically provides enough publicity for the variation's popularity to surge for a while, which is what happened here as well, even though White's achievements cannot be called impressive. After the common obligatory moves - 1 2 . . . .l:.xf7 1 3 fxg4 .l:.xfl + 14 �xfl - Black has several different ways to generate counterplay:

b1) With 14 . . . cxd4 15 cxd4 e5, Black seizes the opportunity to secure his bishop's future. After 1 6 d5 lLlc4, not good for White is 1 7 i.f2 'ili'f6 1 8 �g1 .l:.f8 19 'i!i'e1 i.h6 20 lLlg3 'i!i'a6 with subsequent seizure of the c-file and an attack on the a2-pawn, so White chooses 1 7 'i!i'd3 lLlxe3+ 1 8 'i!i'xe3 'i!i'h4 1 9 h3 i.h6 20 'i!i'd3 .l:.f8+ 2 1 �gl 'ili'f2+ 22 �h 1 , but in the sharp game that ensues Black has sufficient chances as well: 22 . . . 'i!i'e3 23 'i!i'c4 b5 24 'ii'xb5 l:tf2 25 'i!i'e8+ i.f8, Kramnik-Kasparov, Linares 1 999. The preliminary 16 �g1 promises little, as in the endgame after 16 . . . exd4 17 i.xd4 .l:.c8 1 8 i.xg7 �xg7 the extra pawn doesn't qualify as a combat unit of full value, though White does retain a slight initiative.

CLASSICAL EXCHANGE VARIATION: WHITE PLAYS fbe2 45

b2) With 1 4 . . . 1li'd6, Black doesn't object to the building of the pawn-wall to contain the g7-bishop and even provokes White into doing it with gain of tempo, planning to make hay on the weakened light squares: 15 e5 'ii'd5 16 g5 (the immediate 16 ..tf2 is also played) 16 . . . 'ii'e4 1 7 ..tf2 .:f8 1 8 ti:Jg1 (the knight aims for f3, the ideal post in such positions) 1 8 . . . ti:Jc4 1 9 ti:Jf3 ti:Je3+ 20 ..txe3 'ii'xe3 and White has problems keeping his pawn-chain in the centre intact. If he prefers 15 �g1 , then in the endgame after 1 5 .. . 'ii'e6 16 'ii'd3 'ii'c4 1 7 'ii'xc4+ o!Dxc4 1 8 ..tg5 e5 Black has good equalizing chances as well.

b3) 14 . . . 'ii'd7 is a rare but interesting continuation. After 1 5 h3 o!Dc4 16 ..tf2 cxd4 1 7 cxd4 e5, instead of the natural 1 8 'ii'd3 o r 1 8 d5, the mistaken 1 8 dxe5? led to a catastrophe in Van Wely-Sutovsky, European Clubs Cup, Rethymnon 2003: 1 8 . . . ti:Jd2+ 1 9 �el ( 1 9 �gl loses the exchange to 19 . . . ..txe5 20 .:c 1 ..tb2 -the rook cannot leave the first rank) 19 . . . ..th6 20 ti:Jd4 ltf8 2 1 'ii'e2 'ii'c7, and the active black pieces chased the white king upon his return to the centre.

12 ... cxd4 13 cxd4 ..te6 (D)

w

We have reached a key position in the Classical Exchange. Black has generated maximum activity by playing in the most direct and aggressive manner against White's centre. Black can view the c4-square as a prospective trophy, and if he plants a piece there, he can support it with the move . . . b5, which also sets his pawnmajority in motion. In certain cases, White can try to show a black pawn on b5 to be a weakness, but it is undesirable for him to allow the disputed point to be occupied for free and at once. He has three methods to try to prevent Black from settling down on c4 immediately.

14 d5 This is the most radical method. The ex

change sacrifice has been known for a long time; its first flash of popularity flared in the middle of last century and then faded, leaving the variation with an indeterminate reputation. The resources of computer analysis breathed new life into this continuation.

The two alternatives are: a) 14 'ii'a4 presents Black with tempi and al

lows him to create counterplay: 14 . . . a6 1 5 d5 b5 1 6 'ii'b4 ..txal 17 .:xal ..td7 1 8 'ii'd4 f6 (returning the exchange by 1 8 . . . ltc8 19 ..th6 f6 with active operations down the c-file doesn't look bad either) 19 e5 fxe5 20 'ii'xe5 'ii'b8 21 'ii'xe7 .:e8 and Black's chances are at least no worse.

b) 14 :c 1 sacrifices a pawn, and in spite of the apparent danger, accepting it by 14 . . . ..txa2 (D) is the correct course - Black must boldly seize upon any drawback to White's play, since otherwise his strategic aims will have been thwarted for free. Then:

w

b 1) The attempt to cut off the bishop's retreat by 15 d5 allows the freeing advance 1 5 . . . ..tb3 16 'i!Ve 1 (or 16 'ii'd2) 16 . . . e6.

b2) 15 'ii'a4 is the main continuation. Now the bishop can return to its base by 15 . . . ..te6 16 d5 ..td7 17 'ii'b4 e6 1 8 ti:Jc3 exd5 19 o!Dxd5 ..te6 20 .:fd l , when White's powerful central formation and Black's difficulties with the return of the aS-knight compensate for the missing pawn, but not more than that. Leaving the bishop behind the enemy lines for diversion purposes is an interesting idea as well: 15 ... ..tb3 (a move popularized by Anand) 1 6 'ii'b4 b6. White's pieces are placed quite actively, but Black doesn't merely have an extra pawn, but his majority has


now been transformed into a pair of passed pawns. For the sake of setting these pawns in motion, he can acquiesce to considerable concessions in other sectors of the battlefield. The offer of an exchange of queens by either 17 d5 'it'd6 or 17 i.g5 f6 18 i.h4 'it'd6 promises Black a fully.adequate game.

14 ... i.xa1 15 'it'xa1 f6 (D)

w

It is evident that White has at least some compensation: his pieces are more active, and the black king's position is weakened by the absence of the dark-squared bishop. The extent of this compensation is the whole question, as Black has resources of his own. The position is open, and Black's counterplay along the c-file can bring welcome simplification, and the transfer of the queen via b6 with gain of tempo may prove an important resource.

16 i.h6 Every move one could think of has been tried

here - the author counted 17, but the text-move and 16 .l:tb1 , eyeing both the b7-pawn and the b6-square, have been recognized as the main continuations. In reply to the latter a clear-cut path to equality was demonstrated in the game Nayer-V.Mikhalevski, Moscow 2002: 16 . . . i.f7 (by returning the exchange, Black forces a draw) 1 7 i.h6 e6 1 8 i.xf8 �xf8 19 liJf4 exd5 20 exd5 g5 2 1 ltJe6+ i.xe6 22 dxe6 'it'xd3 and White didn't find anything better than perpetual check.

16 ••• .l:te8 17 'it'd4 The interest in the variation was first re

newed with the resurrection of the line 17 �h1 .l:tc8 (17 . . . i.d7? immediately is a mistake, as the knight then changes direction and goes via g3 to e4 to attack f6) 1 8 ltJf4 i.d7 1 9 e5 ltJc4 20 e6 i.a4 21 ltJxg6 hxg6 22 i.xg6 (D), known

since Bannik-Novotelnov, USSR Team Ch, Tbilisi 195 1 , but rejected by theory because of the result of that game.

B

The heavy use of machine analysis gave White several spectacular wins, but everybody has a computer, and ways to defend successfully have been discovered. One is available immediately - 22 . . . �h8 23 'ii'e1 .l:tg8 24 i.e4 i.e8 25 'ii'h4 .l:tg6. Another is calculated through the wild complications arising after 22 . . . ltJe5 23 i.e4 i.c2 24 i.xc2 .l:txc2 25 'ii'd l and then 25 . . . �h7 26 f4 �xh6 27 fxe5 'it'c7 (or27 ... 'it'c8). Those who don't mind a degree of risk can play for a win in the sharp endgame arising after 25 . . . 'it'c7 26 f4 .l:tc1 27 'ii'xc l 'it'xc 1 28 l:txc1 ltJg4 29 h3 ltJxh6 30 g4 .l:td8 3 1 l:tc7 'ito>f8 32 .l:td7 ltJf7.

17 ... i.f7 18 i.b5 e5 19 'ii'f2 .l:te7 (D)

w

This plan is currently in development. The position has been known since the game Gligoric-Portisch, Nice Olympiad 1974. There White played 20 i.e3, lost the game, and, as the result, the idea was forgotten. Today it has been revived, although now the treatment is different White endeavours to clear the long dark-squared

CLASSICAL EXCHANGE VARIATION: WHITE PLAYS tiJe2 47

diagonal and not to withdraw the bishop but, on the contrary, to make use of its presence on h6 as a constant mating threat.

20 f4 White expresses his willingness to exchange

the queens, after which the position still retains middlegame characteristics, as there are still many pieces on the board, and the initiative and attacking chances remain important factors. The e7-rook is not likely to break out into the open, so the exchange will probably have to be returned; the question is whether Black will manage to extract profit from that or the passed d-pawn will assure White an advantage.

A more recent instalment of the discussion in Topalov-Shirov, Morelia/Linares 2008 tested a new idea. In our game the bishop on b5 found itself in a vulnerable position, so White withdrew it at once with 20 .id3, challenging Black to make use of the tempo he is presented with and improve the defence: 20 . . . l:tc8 21 f4 l2Jc4 22 fxe5 l2Jxe5 23 'ii'xf6 .ixd5 24 exd5 l2Jg4 (here 24 . . . l2Jxd3?? is impossible because of 25 d6) 25 'iVf4 lDxh6 26 'iVxh6, and now Shirov should have tried 26 . . . 'ii'xd5.

20 •.• exf4 There are other continuations as well, but

for now this is the possibility that has been studied. Black goes in for a concrete tactical operation.

21 'ii'xf4 'ii'b6+ 22 �h1 .ixd5 (D)

w

Taking immediately on b5 is not an option, since from f6 the queen would threaten a checkmate on either g7 or f8. The counter-offer of the bishop provides g7 with a defender, gaining time to cover f8, so Black succeeds in simplifying the position by exchanging off a passive piece for a more active one.

23 exdS 'ii'xb5 24 'ii'xf6 'ii'e8 25 'ii'd4 White has opened the long diagonal and tied

down the enemy pieces to the defence of the possible mating squares. However, his own king is in need of support, so a tempo has to be spent to prevent trouble on the back rank, and White has to acquiesce to the exchange of his rook, which cannot secure a good square in the face of the quantitative superiority of the enemy major pieces. The last reserve is the knight, which White would love to manoeuvre to f6. This idea was successful in Topalov-Shirov, Wijk aan Zee 2007: 25 . . . l:td8 26 h3 l:tf7 27 l:txf7 'ii'xf7 28 'ii'c3 b6 29 lDg3 and after 29 . . . lDb7 30 l2Je4 White emerged victorious. In a later encounter Rakhmanov-Danin, Russian Team Ch, Sochi 2007, Black managed to defend by 29 . . . l:te8 30 'ii'd4 g5, although one must think Aronian was prepared for this. So Shirov decided to deviate first:

25 ••. l:.f7 This move offers Black another defensive

plan. 26 l:txf7 'ii'xf7 27 il.d2 'ii'g7 (D)

w

This is the point. After 28 'ii'xg7+ <l;;xg7 29 .ixa5 l:tc8, while the material advantage is formally with White, the lack of coordination between White's pieces forces him to treat Black's chances with due respect. Considering the match situation (this was the penultimate game, with Aronian a point ahead), White decided not to look for the needle in a haystack, especially as there may be no needle.

28 'ii'e4 'ii'f7 29 .ic3 l:tc8 30 h3 lDc4 31 'ii'd4 <t;;rs 32 'ii'hS+

If White wishes to continue the struggle, 32 lDf4 is the way.

32 ••• 'ii'g8 33 �d4 'ikf7 34 'ikh8+ 112-lfl


Game 1 0 [087]

Mikhail Kobaliya - Alexander Areshchenko Isle of Man 2006

1 d4 lL)f6 2 c4 g6 3 l'Llc3 d5 4 cxd5 tt::lxd5 5 e4 tt::lxc3 6 bxc3 i..g7 7 i..c4 c5 8 l'Lle2 tt::lc6 9 i..e3 0-0 10 0-0 (D)

B

In this game we shall examine yet another method of mobilizing the black forces. Black completes his development and prepares an advance on the queenside, where he has a pawnmajority.

10 ... .td7 This is the modem preference when Black

opts for this plan. In Fischer's day, 10 .. .'ii'c7 1 1 :c 1 :d8 (D) predominated.

w

This creates some discomfort for the white queen, but there are no concrete threats as of yet, and meanwhile the departure of the rook weakens the kingside defence, and f7 in particular. White must choose the direction of the main advance. Pushing the f-pawn suggests itself, and initially that was the principal plan,

carried out either immediately with 1 2 f4, allowing the development of the enemy bishop to g4 with a temporary pin on the e2-knight, or after the preliminary 1 2 h3, 1 2 'ii'e 1 or 1 2 'ii'a4. This last move is not a signal for the active operations on the queenside, as White lacks the advantage in resources in this sector of the board. Eventually the queen ends up on the attractive post b2, where it takes part in the support of the centre, is safe from attacks and sufficiently mobile for a quick transfer elsewhere. However, his trip has taken a lot of time. Black can meet White's f4 advance with a light-squared blockade according to the following pattern: . . . b6, . . . i..b7, . . . e6 and . . . f5; gradually it became clear that he succeeds in this and obtains rich counterplay.

White then embarked on the search for an advantage with play in the centre, at the same time keeping open the option of switching attention to the enemy king at a convenient opportunity. First we shall examine 1 2 'ii'd2 (D).

B

The immediate raid 12 . . . tt::le5 1 3 i..b3 tt::lg4 14 i..f4 e5 15 i.g3 i..h6 16 i..xe5 ! favours White, but there are also solid continuations such as 12 ... a6, 12 . . . b6 and 12 . . . 'ii'a5. Let's dwell a little on the last one: as the d4-pawn is in jeopardy, White has to choose between 1 3 'ii'b2 and 1 3 :fd 1 . Here too, the time spent decentralizing the queen gives Black a good game after 1 3 'ii'b2 b6. The rook ·move 1 3 .l:lfd l , reinforcing the centre and hoping for an advantage in case

CLASSICAL EXCHANGE VARIATION: WHITE PLAYS lDe2 49

of the queen exchange, looks more logical. At this point tactical complications would be unfounded: 1 3 . . . �g4 14 f3 lt:Je5 (the attempt to reverse the move-order by 14 . . . �xf3 15 gxf3 lt:Je5 doesn't work, as after 17 dxe5 the compensation for the queen is more than sufficient) 15 �d5. Therefore Black gets on with his development with 1 3 . . . �d7. White has fully mobilized, but trying to throw himself on the enemy king leads only to a perpetual check: 14 �h6 cxd4 1 5 �xg7 �xg7, and now the endgame is quite acceptable for Black, so 16 �f4 dxc3 17 �xf7+ �h8 1 8 .l:td5 (it's the endgame again after 1 8 lt:Jxc3 .l:tf8) 18 . . . �b4 19 lt:Jf4 �xc4 20 lt:Jxg6+ hxg6 21 �xg6 (2 1 .l:th5+ wins the queen, but Black again has fully adequate compensation) 2 1 . . .�g4 or 2l . . .e5.

Today White's hopes for an advantage are connected with 12 �f4 (D).

B

The counter-stab 12 . . . e5?! appears dubious; White continues with 1 3 �g5 .l:td6 14 �d5 and in reply to Black's unnatural manoeuvres he simply improves the placement of his pieces.

In the vast majority of games Black picks l 2 . . . �d7, for the time being blocking his own bishop's diagonal but forcing White to clarify the situation in the centre. Then 1 3 dxc5 is occasionally seen. In agreeing to the destruction of his centre, White temporarily acquires an extra, even if doubled, pawn, but his main achievements are the advantage in space and active pieces. Specifically, the d5-bishop can be driven away with . . . e6, but then the other bishop will settle on d6. The main continuation is l 3 . . . �e8 14 �d5 �d7 15 �d3, and if 15 . . . e6, then White can seek an advantage with both 16 �b3 and 16 �xc6. However, the most important continuation is 1 3 d5 lt:Ja5 14 �d3. Black often chooses

l 4 .. . e5 here, closing the position and locking in the enemy light-squared bishop; also the e5-pawn takes over control of the d4-square from its colleague on c5, freeing the latter to advance. Still, in the resulting positions the sum of White's positional pluses outweighs the sum of the drawbacks. Another option is 14 . . . b5 15 .l:tb l a6, when the search for an advantage continues with 16 �c l or 1 6 �e3.

That is why Black started wondering if he wasn't determining the queen's position prematurely, and 10 . . . �d7 (D), to which we now return, came to the fore:

w

On d7, the bishop supports . . . b5 and vacates the c8-square for the queen.

ll .l:tbl This move is mostly a prophylactic measure

against . . . b5. At the moment the b7-pawn is indirectly defended by the possibility of . . . lt:Ja5.

l l ... �c7 Why not to c8 at once? It is useful to lure the

bishop to f4, where it will come under attack after . . . e5, and the defence of the d4-pawn is weakened. Also possible is l l . . . a6, renewing the threat of the advance of the b-pawn. Then White tries to make use of the weakening of the dark squares by means of 12 dxc5 with subsequent pressure on the b7-pawn, while Black hopes to take advantage of the shattered pawn-centre. Here is a characteristic example: 12 . . . lt:Ja5 13 �d3 'flc7 14 .l:tb6 .l:tfd8 15 �bl e6 16 �g5 .l:te8 and now the thrust 17 c6 gave White an advantage in Kaidanov-Nakamura, USA Ch, San Diego 2004. The text-move threatens the bishop on c4.

12 �f4 If White withdraws the bishop at once with

12 �d3, to avoid being forced to advance in the


centre before his development is complete, then Black increases the pressure by placing one of the rooks on d8. We shall illustrate the options w for both sides with the game Sargissian-Lupulescu, European Team Ch, Khersonisos 2007: 12 .. .l:lad8 1 3 'ili'c2 lba5 14 dxc5 ..i.e6 1 5 c4 lbc6 16 lbf4 ..i.c8 17 lbd5 'ili'd7 1 8 f4 e6 and now White left the knight en prise, giving it up for two centre pawns by 19 :fd I .

12 .•• 'i!i'c8 13 d5 lba5 1 3 . . . lbe5 gives the white centre an opportu-

nity for a concerted advance: 14 ..i.xe5 ..i.xe5 1 5 f4 ..i.g7 16 e5 . Perhaps this i s not lethal, but the black king cannot feel safe.

14 ..i.d3 e5 (D)

w

This move is a typical idea in this formation. It is clear that the en passant capture doesn't come into consideration. Black leaves his opponent with a protected passed pawn, but this is counterbalanced by several pluses of his own: • White's pawn-majority in the centre becomes

static, whereas Black's three-to-two queenside advantage is quite dynamic;

• The white pieces' active potential has significantly decreased;

• The c5-pawn is freed from its duty guarding the d4-square and can now advance. 15 ..i.d2 White's plan is sure to include f4, so 15 ..i.g3

is more common, in order to maintain the attack on the e5-pawn. However, this gives Black a chance to embark upon a very sharp course: 15 . . . f5 16 f4 fxe4 17 ..i.xe4 lbc4 1 8 �d3 ( 1 8 d6! ?) 1 8 . . . lbd6 19 fxe5 lbxe4 20 � xe4 ..i.f5 2 1 �c4 ..i.xb1, Carlsen-Navara, Wijk aan Zee 2007.

15 ... c4 16 ..i.c2 b5 (D) Not only setting the pawns in motion, but

also clearing the retreat path for the knight.

White can get rid of one potential weakness by playing 17 a4, but at the moment an open a-file would be in Black's hands.

17 f4 f6 Maintaining the pawn on e5 . The alternative

is to settle for piece control over that square, but then both sides' pieces are activated. If only the knight could be quickly relocated to e5 . . .

18 ..i.e3 After 1 8 fxe5, the zwischenzug 1 8 . . . 'i!i'c5+

hands the initiative to Black. 18 .•• lbb7 19 'ili'd2 lbd6 20 �h1 White is in no hurry to exchange on e5 be

cause he hopes to profit from keeping the option of playing f5, and also to avoid simplifications along the f-file. With this move he clears the knight's path to the active post on f3.

20 ••• f5 (D)

w

However, White's slow play and slightly sensitive back rank allow Black to conjure up some threats, and his fully mobilized black army begins counterplay in the centre.

21 lbg3 There are many ways in which a series of ex

changes can take place, and White chooses one that does not activate Black's pieces, but he

CLASSICAL EXCHANGE VARIATION: WHITE PLAYS tbe2 51

relaxes his control over the c3-square and underestimates Black's chances. Although it was about time to say goodbye to the hopes for an w advantage, there were several ways to maintain equality, such as 2 1 fxe5 i.xe5 22 i.d4. Taking the pawn is risky: after 2 1 exf5 lDxf5 22 i.xf5 i.xf5 23 .l:txb5 i.d3 the bishop is immensely strong, and it is not so easy for the b5-rook to re-turn, which aggravates the back-rank problems.

21...exf4 22 i.xf4 lDxe4 (D)

The endgame after 27 i.e3 (or 27 i.h6) w 27 . . . bxc3 28 .l:txf5 cxd2 29 .l:.xf8+ .l:txf8 30

23 lDxe4?! It is more prudent to keep the knight by 23

i.xe4 fxe4 24 lbxe4, but after 24 . . . .tf5 25 lbd6 1i'd7 it transpires that the voyage to d6 presents Black with additional possibilities, and in reply to 26 lDxf5?! Black now has the recapture with the rook, immediately attacking the d5-pawn, as an extra option. Therefore White must fight for a draw with 26 llxb5 i.d3 27 lle1 llxf4 28 1i'xf4 i.f8 29 llb7 i.xd6 30 .l:txd7 i.xf4 3 1 .:l.ee7.

23 .•• fxe4 24 .txe4 .tr5 25 i.xf5 1Wxf5 26 'it>gl ?!

Had White realized the full extent of the impending danger, he would have probably started seeking chances with 26 g3 or 26 h3.

26 ... b4 (D) 27 :Xb4 27 i.d6? loses at once: 27 .. . bxc3 28 1i'e2 (28

1We3 is met with the simple 28 . . . 1i'xd5) 28 . . . c2 29 .l:.bc I i.d4+ 30 'it>h 1 'iVxfl + 3 1 .:l.xfl .:l.xfl + 32 'iVxfl i.e3 33 i.a3 .:l.b8.

i.xd2 leaves no hopes for salvation either. The c4-pawn is instantly transformed into a shock combat unit.

27 ••• g5 28 .l:txc4 (D)

B

28 •.. .te5 To obtain a third pawn for the piece, White is

forced to expose his king, and Black quickly decides the game with a direct attack.

29 g3 gxf4 30 gxf4 'it>h8 31 'it>bl .l:tad8 32 .:tel .tf6 33 .:l.dl .:l.gS 34 1i'd3 1Wg4 35 1Wn .l:tde8 36 d6 .l:te2 37 h3 1i'g3 0-1

The system of development with the bishop on c4 and the knight on e2 is experiencing a new surge in popularity. This is particularly true of the lines with the exchange sacrifice, examined in Game 9, where important new games appear on a regular basis.

4 Modern Exchange Variation : White Plays ti:Jf3

On the basis of games from the 1920s, the development of the knight to f3 was for many decades considered inferior, because the knight is subject to pin and possible exchange by . . . i.g4, and thus cannot play a full part in the struggle for d4. To put it another way, it allows Black to use his lightsquared bishop in the fight for the dark squares in a way that was not available in the lines we examined in the previous chapter. White has a similar resource in that . . . ltJc6 can be met by i.b5, but this is less potent since an exchange on c6 brings another pawn to the c-file which can be used to fight for the d4-square.

We should consider the question of why White is so preoccupied with maintaining a pawn on d4. After all, a game cannot be won by merely reinforcing the centre and admiring it from a distance. It is only natural that sooner or later the pawns must advance, but for the advance to be effective it must be supported by a fully mobilized army, and it is desirable to keep open the option of beginning the advance with either centre pawn. And if Black does manage to force the premature advance of one of them, it would be nice to obtain compensation in the form of other positional or material concessions.

The rebirth of the system began around thirty years ago and it fairly quickly gained a dominant place in modem practice. One of the key points is that at the cost of allowing . . . i.g4 ideas, White is able to develop his pieces to more natural squares where they are better able to support the advance of the pawns. And the . . . i.g4 idea itself turns out to have its dark side, as it leaves the b7-pawn unguarded, a factor that is highlighted by the move l:tb 1 . This idea is important in many of the ltJf3 lines at some stage, while the main line with an immediate l:tbl is discussed in the next chapter. This short chapter forms an introduction to the ltJf3 Exchange and its major ideas, while also examining two other development plans for White. Note that in many cases the move l:tc l is important, with ideas of penetrating to c7 if Black opens the c-file, or as a preparation for the d5 advance if Black instead maintains the tension.

Game 11 (Nayer-Khamrakulov) is devoted to the most natural placement of bishops on e2 and e3. White castles quickly and is ready to drive the c6-knight away by playing d5 .

In Game 12 (Vitiugov-Zhao Jun) White immediately develops his queenside pieces, removing all possible objects of attack from the firing-line of the g7-bishop. White meets the black queen's sortie with a standard pawn sacrifice.

Game 1 1 [085]

Evgeny Nayer - lbragim Khamrakulov Ubeda 2001

1 d4 ltJf6 2 ltJf3 g6 3 c4 i.g7 4 ltJc3 d5 5 cxd5 ltJxd5 6 e4 ltJxc3 7 bxc3 c5 (D)

Let's examine the prophylactic continuation 8 h3. Despite the loss of a tempo, this move enables White to hold the centre for a while. After 8 i.e2

MODERN EXCHANGE VARIATION: WHITE PLAYS ll:Jf3 53

w

8 . . . 0-0 9 �e2 Black has two reasonable options:

a) 9 . . . "Llc6 10 �e3 cxd4 1 1 cxd4 1Va5+ 12 �d2 1Va3 and now 1 3 d5 is forced if White does not wish to repeat moves or give up the right to castle. Black rarely accepts the exchange sacrifice; compared to Game 9, the differences favour White, but after 1 3 . . . lt::le5 (or 1 3 . . . "Lld4) 14 "Llxe5 �xe5, 1 5 0-0 �d7 gives Black his fair share of the play. If the offer is limited to a pawn, it should be taken - 1 5 .l:!.bl 1Vxa2; the potential value of the connected passed pawns constitutes a reserve that allows dumping some material if necessary.

b) The search for a set-up where the move h3 looks the least useful leads to the selection of the fianchetto of the light-squared bishop 9 . . . b6 - with standard play and practically an extra tempo.

If the right to move is conferred to the flbishop, it has two reasonable options, one of them the text-move, and the second 8 �b5+, when Black has two main replies:

a) 8 ... �d7 9 �xd7+ (in case of9 �e2 Black often chooses the aforementioned plan with the transfer of the bishop to c6) 9 . . . 1Vxd7 10 0-0 0-0 1 1 �e3 cxd4 12 cxd4 "Llc6. The bishop exchange has simplified the position, and Black has no problems finding good posts for the remaining minor pieces, while the open file facilitates the reduction of the number of major pieces.

b) 8 . . . "Llc6. Now 9 d5 1Va5 1 0 .l:!.bl a6 fails to disturb Black, while the endgame after the relatively forced sequence 9 0-0 cxd4 10 cxd4 0-0 1 1 �e3 �g4 1 2 �xc6 bxc6 1 3 .l:!.c l 1Va5 14 1fd2 1fxd2 1 5 "Llxd2 .l:!.fd8 16 lt::lb3 a5 is not considered to be problematic for him.

8 •.. 0-0

The immediate inclusion of another piece in the attack upon d4 by 8 . . . lLlc6! gives White the opportunity to offer yet another version of the exchange sacrifice: 9 d5 �xc3+ 10 �d2 �xa1 1 1 1fxal lLld4 1 2 lLlxd4 cxd4 1 3 1Vxd4 (D).

B

Nevertheless, Black's chances are not worse whether he intends to keep the gift with 1 3 . . . f6 or agrees to return it after 1 3 . . . 0-0 14 �h6 f6. The attempt to hold the centre by 9 �e3 �g4 forces White to play 10 e5, which can be considered a positional concession.

Of course, having read the introduction to this chapter, you will be wondering what happens after 8 . . . .ig4, a thematic way to increase the pressure on the centre. Then White can strive for a minimal advantage based on the bishop-pair: 9 .l:!.b1 (this move should not come as a surprise either) 9 . . . lt::lc6 10 .l:!.xb7 0-0 1 1 0-0 �xf3 12 �xf3 cxd4 1 3 cxd4 lt::lxd4 14 �g5.

9 0-0 cxd4 If he wishes to avoid mass exchanges, Black

can prefer 9 . . . b6 here as well, even though this gives White a respite from the defence in the centre for one move, and allows him to arrange his forces more actively: 10 .ig5 .ib7 1 1 1Vd3 1Vd7 12 .l:!.adl and the future passed d-pawn promises White the better chances. Two other standard continuations can transpose, though White acquires the extra option of d5 (at once in reply to 9 . . . lLlc6, in case of 9 . . . �g4 after 10 �e3 lt::lc6). With a pawn remaining on c3, the g7-bishop has nothing to treat itself to on the long diagonal, and this renders White's d5 advance extremely potent. The exchange in the game clears the diagonal.

10 cxd4 lt::lc6 11 �e3 �g4 (D) Black has succeeded in including all his mi

nor pieces in the attack on the d4-pawn, and its


w

defence has no reserves left. White must either advance one of the pawns or counterattack b7.

12 d5 This thematic response is also best. 12 e5 locks the g7 -bishop out of the game for

a while, but the price seems exorbitant - the light squares are weakened, and the centre is deprived of the prospects of further advance. Black's hopes of wresting the initiative from White appear well-founded, although the d4-e5 pawn-wall is hard to breach. It is also not clear what advantage there is to hope for in the line 12 l:!.bl ..ixf3 1 3 .txf3 ..ixd4.

12 ••. lt:le5 White's previous move doesn't really sacri

fice the exchange, as after 12 . . . -txal 1 3 �xal ..ixf3 14 ..ixf3 lt:la5 15 ..ih6 f6 he can win it back at once. However, the position is simplified too much then, so more interesting is 1 6 ..ig4 �d6 17 ..ie6+ 'it>h8 1 8 h4 lt:lc4 19 �c3 b5 20 �h3 with a strong initiative, MastrovasilisChuchelov, European Ch, Warsaw 2005.

13 lt:lxe5 ..ixe2 14 �xe2 ..ixeS (D)

w

Both sides have played logically and consistently, but White has achieved his main aims. Despite the significant simplification, he retains

a certain initiative, and his central superiority looks a more weighty factor than Black's twoto-one queenside preponderance. Moreover, the black king' s cover is weakened, especially if the bishops are exchanged. However, if White fails to make use of his dynamic trumps, the assessment will change.

Now White must decide where to place the rook. The immediate occupation of the open file is a routine decision that promises further exchanges, yet each side needs to keep the open line constantly under control in order not to cede it to the opponent. In a situation where there is only one open file, control over it constitutes a serious advantage. The most natural placement of the rooks is on bl and d l or on d l and f l .

15 l:!.adl Here the rook supports the creation of a

passed pawn and prepares the bishop exchange, so this is a logical decision.

Picking 15 l:!.abl creates the opportunity to offer a queen exchange on b5 after 1 5 . . . �d7, while 1 5 ... b6 weakens the c6-square, and White will try to use it to seize the c-file. After an exchange of queens, the passed pawn may prove more dangerous on c6 than on the d-file.

lS ••• �aS (D) Along with 15 . . . �d7, this is the most natu

ral reply. Black connects rooks and trains his queen's sights on the enemy pawns on a2 and d5, several squares on the c-file and, as a prophylactic measure, protects a7, freeing the rook from that duty.

w

16 h4 The idea behind this move is clear - the h

pawn is destined either to weaken the enemy king's cover or, should it be allowed to reach

MODERN EXCHANGE VARIATION: WHITE PLAYS l:iJj3 55

h6, to take direct part in the attack. However, White's follow-up doesn't create the impression that these are elements of the same plan. 16 f4 i..g7 posed more problems, with the choice between 17 f5, 17 e5 or 17 'ii'f2, tying one rook to the defence of the a7-pawn and the other to the f-file.

16 ... l:tac8 17 i..d4 'ii'c7 Taking advantage of the opportunity to seize

the file or centralize the queen. Black agrees to allow the creation of a passed pawn on the dfile after the exchange of the queens, as long as that is not accompanied with an attack on his king.

18 'ii'e3 i..xd4 19 'ifxd4 (D)

B

19 ••• 'ii'c5 Now White manages to neutralize Black's

control of the open file while keeping the queens on the board. 19 . . . b6 is more accurate.

20 'ifb2 b6 21 l:tcl 'ifd6 22 g3 Nevertheless, White has nothing concrete,

so decides to maintain the position largely unchanged, hoping for an inaccuracy on Black's part. The primary purpose of the text-move is to deny f4 to the black queen, while at the same time consolidating the kingside pawn-chain.

22 •.• f6 23 �g2?! Stepping into the line of a possible capture

with check on d5 is not the best idea. 23 ••• 'ife5?! Inexplicably, Black voluntarily spoils his own

pawn-structure. He could have taken the favourable opportunity to play 23 . . . f5 24 f3 e6. After 25 l:txc8 l:txc8, 26 l:td 1 fxe4 27 fxe4 exd5 leaves the white king as exposed as his black counterpart, while 26 'iff6 fxe4 hands Black the initiative.

24 'ii'xe5 fxe5 (D)

w

It is possible that there are real drawing prospects in this endgame too. However, the number of black weaknesses has increased, while his chances for counterplay have diminished.

25 l:tfd1 �f7 26 f4?! In his hurry White helps his opponent get rid

of the weakness on e5 . It would be more expedient to bring the king into the centre first and only then start active operations.

26 .•. exf4 27 gxf4 l:tfd8 (D) There are good drawing chances in the line

27 . . . e6 28 d6 l:txc1 29 l:txc l e5 30 l:tc7+ �e6 31 fxe5 �xe5 (but not 3 1 . . . l:tf7? 32 �g3, as all pawn endgames are won for White).

w

28 �f3 �f6 29 �e3 e5?! Why help White to create a pair of passed

pawns? White has no clear-cut plan of realization of his positional advantage, and in the coming play the drawing tendencies characteristic of rook endgames leave Black some hope for salvation. For the moment Black could adopt a waiting policy, for example with 29 . . . �f7.

30 :n h6 31 �d3 White covers the c-file and, after exchanging

on c8 and e5, plans to drive the king away from


the blockading square with the other rook (infiltrating via f7 or after gaining control of the f5-square with the help of the h5 advance) and w set the pawn-pair in motion. Therefore Black himself takes on f4.

31 .•• exf4 (D) 32 <;td4? White returns the favour. There was an easy

win after 32 .l:txf4+ <;te5 and now 33 .l:txc8 .l:txc8 34 .l:tf7 or 33 .l:tcfl .

32 •.• .l:txcl? This should settle the result, whereas the im

mediate 32 . . . g5 would level both sides' chances with a probable draw: 33 e5+ <;tf5 34 hxg5 hxg5 35 .l:txc8 .l:txc8 36 e6 g4 37 e7 f3 38 .l:tel g3 39 d6 f2 40 .l:te5+ and one of the kings will succumb to a perpetual check.

33 .l:txcl g5 34 .l:tc6+? Now 34 . . . <;tg7 ! would salvage a draw; 34

e5+! wins. 34 ••• <;tf7? 35 e5 <;tg7 36 h5 f3 37 .l:tg6+ <;th7

38 e6 g4 39 l:txg4 .l:tf8 40 e7 1-0

Game 1 2 [085]

Nikita Vitiugov - Zhao Jun World Junior Ch, Erevan 2006

1 d4 lt:'lf6 2 c4 g6 3 lt:Jc3 d5 4 cxd5 lt:Jxd5 5 e4 lt:Jxc3 6 bxc3 iJ..g7 7 lt:'lf3

The desire to develop the queenside first is often fulfilled with the immediate 7 iJ..e3 c5 8 �d2 (D).

B

In most games this comes down to a transposition of moves, but we should note the independent continuation 8 .. .'�a5 9 .l:tbl . White wants to force the queen exchange early, while Black hasn't castled and the king's rook is far from the events in the centre, plus after the exchange the bl -rook controls the b4-square and attacks the pawn on b7. If Black obliges by 9 . . . cxd4 10 cxd4 �xd2+ 1 1 <;txd2 0-0. then

White can play 1 2 iJ..d3, when instead of retreating to the first rank the king has a post on e2 ready for him. However, there are problems with the centre here as well: 12 . . . e6 13 lt:Jf3 .l:td8 14 .l:thc l lt:'lc6 and now 15 .l:tc4 looks clumsy, while 1 5 e5 f5 (but not 15 . . . f6 1 6 <;te2) gives Black counterplay. Instead of the exchange, Black can play 9 . . . b6 or 9 . . . a6, when White usually reckons that he has succeeded in weakening Black's position and moves the rook to c I . This move-order also grants the d4-pawn the freedom to advance, and the c-pawn can later support its neighbour, but building the pawn wedge c4-d5-e4 takes time (unless Black helps by developing the knight to c6) and hasn't received general acclaim. Black gets ahead in development and successfully counterattacks the centre with the undermining . . . e6 and . . . f5 .

If Black doesn't reply with concrete threats, then White can initiate the bishop exchange on h6.

7 ••• c5 8 iJ..e3 (D) The plan is l:tc 1 , clearing the long diagonal

of potential targets and securing the c3-pawn. This grants White more freedom in the centre but delays castling.

8 ... �a5

MODERN EXCHANGE VARIATION: WHITE PlAYS tiJj3 57

B

This is the main line. Let's examine nonstandard methods of counterplay applicable to this move-order:

a) 8 . . . ..ig4 is of course thematic, but the development of the bishop before castling allows the possibility of checks from a4 or b5, though neither achieves much. The pattern of play after 9 ..ib5+ will be familiar to us - 9 . . . ltJc6 10 0-0 cxd4 I I cxd4 0-0 I 2 ..ixc6 bxc6, when the bishop-pair balances the weakness on c6. Also little is promised by 9 'ii'a4+ ltJc6 (after 9 . . . ..id7 10 'ii'a3 the queen's new post can be considered advantageous) 10 ltJe5, when Black can choose 10 . . . cxd4 I I ltJxc6 bxc6 I2 cxd4, or he can play more sharply with 10 . . . ..ixe5 I I dxe5 'ii'c7. White retains hopes for an advantage in the line 9 'ii'd2 cxd4 10 cxd4 ltJc6 I I :.d I ; in accepting a defect in his pawn-structure, White reckons that the bishop-pair and the pawn-centre (and a strong passed d-pawn in the future) will prove more important.

b) 8 . . . ltJc6 can lead to original play after 9 l:.cl :

bi) 9 ... 'ii'a5 is met by 10 d5, when 10 .. . ..ixc3+ I I :xc3 'ii'xc3+ I2 ..id2 'ii'f6 13 dxc6 'ii'xc6 gives White the better chances, so the knight is forced to move to e5 or even b8.

b2) The preliminary exchange 9 . . . cxd4 10 cxd4 'ii'a5+ can be met with a pawn sacrifice, viz. I I ..id2 'ii'xa2 I 2 d5 ltJe5 I 3 ltJxe5 ..ixe5 14 ..ib5+, when I4 . . . ..id7 is recognized as the lesser evil. However, few players are willing to suffer for a pawn with the king on d7.

c) 8 . . . 0-0 usually transposes to the main lines which we shall examine below, but we should mention the possibility 9 'ii'd2 ..ig4 I 0 ltJg5. Along with the prophylactic h3, another method of neutralizing the pin on the knight is carrying out that same h3 with a simultaneous

attack on the enemy bishop. The knight's position on g5 is shaky, but Black is reluctant to drive it away with pawns, as that creates weaknesses, and in the meantime it takes the e6 retreat-square away from the bishop. White keeps the centre, but Black secures the coveted c4-square: 10 . . . cxd4 I I cxd4 ltJc6 I2 h3 ..id7 I3 :.b I :c8 I 4 ltJf3 ltJa5 15 ..id3 ..ie6 I 6 0-0 ..ic4, Tallaksen-Holmsten, Nordic Ch, Bergen 200 I .

9 ..id2 The development of the bishop from c I to d2

via e3 looks peculiar, but it has its adherents. The potential threats against the queen on a5 provide some justification, and take the question of the queen exchange off the agenda.

9 'ii'd2 (D) is more natural and common.

B

Black can exchange queens by 9 . . . cxd4 IO cxd4 'ii'xd2+, but there is no need to hurry with this until White has defended c3 and is threatening to play d5. With the immediate exchange, the recapture with the knight relocates it to a less vulnerable post on b3, so Black usually picks one of the following three standard continuations:

a) 9 . . . ..ig4 lost its former popularity after Kasparov's defeat against Kramnik. Perhaps the resulting endgame can be held, but painstaking defence is required of Black: I 0 l:.b I a6 I I .l:.xb 7 ..ixf3 I2 gxf3 ltJc6 I 3 ..ic4 0-0 14 0-0 cxd4 I 5 cxd4 ..ixd4 I 6 ..id5 11i'xd2 ( 1 6 . . . ..ic3? ! was played in Kramnik-Kasparov, BGN World Ch (game 2), London 2000) I7 ..ixd2 :ac8, Tregubov-Bacrot, French Team Ch, Montpellier 200 I .

b ) After 9 . . . 0-0 White places the rook on b i or c l :

b i ) 1 0 :bi does not protect the c3-square, and as a result, does not render the d-pawn

58 CHESS EXPlAINED: THE GRUNFELD

mobile. However, in case of the queen exchange the rook is better placed, as it prevents the knight's jump to b4 and forces the bishop to take the attack on b7 into account. In addition, .l:tb5 becomes an immediate threat. Then:

b l l ) A complex struggle arises after the sacrifice of the queen for a rook and a bishop by 10 . . . tbc6 1 1 l:tb5 cxd4 12 .l:txa5 dxe3 1 3 'iVxe3 tbxa5'. While Black does not get a full material equivalent, the weakness of White's queenside grants Black some compensation.

b12) Of course, Black is not obliged to change the material balance so radically and can meet the threat with 10 . . . a6, when the rook can consider its mission accomplished and move on by 1 1 .l:tc 1 , when the d4-pawn is ready to advance. Now if Black exchanges queens, there is an extra option of attacking a d8-rook with .tb6. The prophylactic l l . . . e6 gives White a tempo for the bishop exchange 12 .th6, when the endgame after 1 2 . . . cxd4 13 .txg7 �xg7 14 cxd4 'iVxd2+ 1 5 �xd2 is to his advantage, since the king has acquired a comfortable square on e3.

b2) 10 .l:tc 1 (D) is more common.

B

Then: b21 ) Following the exchange 10 . . . cxd4 1 1

cxd4 'iVxd2+ White has played all possible recaptures on d2. After 12 tbxd2 we should note Kramnik's original idea 12 . . . l:td8 1 3 tbf3 tbc6 14 d5 (on the board is the same position as in the variation that is commonly reached via a different move-order, which we shall examine below - 12 �xd2 tbc6 13 d5 .l:td8 14 �el - with the sole difference that White has kept the right to castle). The popularity of 12 .txd2 has grown; here is a game featuring an interesting tactical struggle: 12 . . . .tg4 13 .te3 tbc6 14 d5 .txf3 15

gxf3 tbd4 1 6 .tg2 l:tfc8 17 �d2 e6 1 8 �d3 exd5 1 9 .txd4 dxe4+ 20 fxe4 .l:.d8 21 e5 .txe5 22 .l:.c4 b5 (Kotanjian-Pantsulaia, Dubai 2007) and now 23 .l:tb4! l:tac8 24 �e4 gives White good winning chances.

b22) Recently the continuation 10 . . . .l:.d8 1 1 d5 e6 (D) has been preferred.

w

At first sight the position after 12 c4 should be to White's advantage - he has a powerful centre, which usually counts for more than an advantage on a flank, while development-wise there is approximate parity. However, practice has demonstrated that Black's chances are not worse thanks to the excellent bishop on g7 and in the future the other one on f5. After 12 . . . 'iVxd2+ the main reply is 1 3 tbxd2 (the bishop keeps c5 under attack and blocks the e-file), when 13 . . . b6 14 .te2 tba6 15 0-0 lbb4 (this transfer of the knight to c3 is the point of the set-up) 16 a3 tba2 17 .l:.cel ( 1 7 .l:tc2 tbc3 is risky for White after 1 8 .tf3 exd5 or 1 8 .td3 .ta6) 17 . . . tbc3 1 8 .tf3 .tb7 1 9 g4 exd5 20 exd5 b5 led to mass exchanges in Zhou Jianchao-Wang Yue, Shandong Zona1 2007. The other recapture, 13 �xd2, hasn't brought White much joy either; even though the queens have disappeared from the board, there are still many pieces left, and the king is uncomfortable in the centre: 1 3 . . . tba6 14 a3 b6 15 �c2 .tb7 16 g4 tbc7 17 .tf4 .l:tac8 1 8 .te2 exd5 1 9 exd5 tbe6, A.Grigoryan-Rodshtein, World Junior Ch, Erevan 2007.

c) 9 . . . tbc6 is the most natural continuation, forcing the recapture with the king after the exchange of queens. White is left with the decision where to post his rook:

c l ) On b l it forestalls the knight leap to b4 and gains a tempo for the development of the fl -bishop: 10 .l:tbl cxd4 1 1 cxd4 'iVxd2+ 1 2

MODERN EXCHANGE VARIATION: WHITE PLAYS CiJj3 59

'iti>xd2 0-0 1 3 dS and now one line is 1 3 . . . l:td8 14 �d3 tDaS, when we see a drawback to the rook being on b 1 : the threat of an invasion on c7 is not an option. The dark-squared bishop can find an active post on either cS, f4 or gS, whereas Black' s counterplay is based on the undermining thrust . . . fS. The knight may also withdraw at once by 13 . . . tiJa5, when the exchange sacrifice 14 �d3 e6 l S �cS :td8 16 �e 7 l:td7 17 d6 a6 1 8 l:tb6 tDc6 19 l:txc6 bxc6 20 l:tb 1 is interesting, though insufficiently tested. We should reiterate that the rook's move to b l does not pose the threat of dS, and Black can transpose into one of the lines examined earlier.

c2) 10 l:tcl is more common here too. Now the exchanging line is 10 ... cxd4 1 1 cxd4 'ifxd2+ 1 2 'iti>xd2 is more forced. 1 2 .. . 0-0 threatens to include yet another unit in the attack on d4, and the need to disturb the c6-knight arises. White can attack it with 1 3 �bS, but Black gets a good game with 1 3 . . . fS. The other method is 1 3 dS; this move must be made at once and not as a reply to . . . l:td8, so as not to allow the immediate counterblow . . . e6. Black plays 13 . . . l:td8 anyway, when the attempt to block the d-file with 14 �d3 allows 14 .. .'�Jb4, and the knight manages to take part into the counterattack on the dS-pawn: l S �bl e6 16 a3 exdS. The main continuation is Kasparov's 14 'iti>el (D), which he introduced in 1 98 1 .

B

This may not look very pretty, but there is no other way to meet . . . e6. The knight usually moves to aS, though after 14 . . . tlJb4 no advantage for White has been demonstrated either. The pair of hard-working bishops compensates for the possible spoiling of the pawn-structure by l S �d2 tDa6 16 �gS 'ifi>f8 17 �xa6 bxa6 1 8 :tc7 :td7. After 14 . . . tlJaS l S i.gS �d7 1 6 �d3,

Black can opt for 16 . . . l:tdc8 17 'iti>e2 (as mentioned above, Kramnik, who arrived at this position via a different move-order and preserved the right to castle, played 1 7 0-0 after all, but didn't achieve anything concrete in KramnikSvidler, Dortmund 2004) 1 7 . . . e6, when 1 8 l:txc8+ l:txc8 1 9 :te l :txc 1 20 �xc 1 exdS 21 exdS bS leads to a minor-piece ending where White's greater piece activity is counterbalanced by Black's queenside pawn-majority. But Black can play the sharper l 6 . . . fS 17 eS �e8 1 8 d6 l:tdc8 with good prospects in the subsequent complex struggle.

9 ••• 0-0 10 �e2 tDc6 A rare move that amplifies the effect of the dS

advance. The common continuation is 10 . . . �g4 1 1 0-0, when l l . . . l:td8 brings up the reserves to mount further pressure on d4. White has at his disposal the sharp pawn sacrifice 12 dS �xf3 1 3 �xf3 �xc3 14 �xc3 'ii'xc3 I S l:tbl b6 16 eS, as in Korchnoi-Balinov, Mitropa Cup, Leipzig 2002. The preliminary 1 1 . . .e6 is more restrained, with good prospects of equality.

11 0-0 l:td8 12 l:tc 1 (D)

B

Playing dS will cost White one of the pawns, and the one on a2 is the less important.

12 ••• cxd4 The immediate 12 . . . 'ii'xa2 is also possible,

when White combines the threat of the advance in the centre with play against the infiltrator in his rear, the enemy queen: 13 dS tDeS 14 tDxeS �xeS l S �gS. By exchanging on d4 first, Black deprives the rook of the al -square but gives the d2-bishop an extra diagonal.

13 cxd4 'ii'xa2 14 d5 tDeS 15 �b4 tDxf3+ 1 S . . . tiJc6 deserved consideration, as there are

no particular achievements for White to be seen after either 1 6 �c4 'ii'b2 1 7 �cS b6 1 8 'ii'a4


bxc5 1 9 �xc6 i.g4, or in the line 16 .l:lc2 �a4 17 i.d2, when 1 7 . . . lZ:le5? loses to 1 8 .:txc8, but 1 7 . . . lZ:ld4 is not bad.

16 i.xf3 .trs (D)

w

17 i.c3 �a6 18 �d4 f6 19 e5 fxe5

22 ... b6 23 d6 l:tb8 24 l:tc7 l:txd6 25 ..tb4 l:td8?!

25 . . . l:td7 26 �c 1 l:td8 27 .l:txa7 i.f5 promised more.

26 i.xe7 .l:te8 27 �e5 i.g7 28 �d6 i.e6 29 i.c6 l:.ec8 30 .:txa7 b5?

This again brings Black to the brink of defeat. 30 . . . �f5 was correct, removing the queen from the gaze of the a7-rook, and forcing White to take measures against . . . i.e5.

31 .:tel .:tb6? (D) A losing move. Some hope remained after

3 l . . . i.c4 32 i.d7 i.c3 33 i.xc8 l:txc8, staking everything on the b-pawn. Still, the adventures are not over yet.

This exposes the king, but 1 9 . . . i.g7 20 d6 w fxe5 21 dxe7 is still more dangerous, since on e7 the pawn creates numerous tactical threats.

20 �xeS �f6 21 �e3 �f7 (D)

w

22 i.a5?! White has a selection of good continuations.

Including the last idle piece in the action by 22 l:He1 is the most solid one, while a tempting idea is to let the rook through to c7 by 22 .tal (or 22 i.b2), considering that 22 ... e6 can be met with 23 i.g4 and now 23 . . . exd5? is no good because of 24 l:txc8 and 25 i.e6. The text -move lets a part of the advantage slip away.

32 i.e8? 32 l:tc7 l:txc7 33 �xc7 l:.a6 34 �d8+ i.f8 35

i.xb5 wins easily. 32 •.. .:r.xe8? The last mistake. There were still some defen

sive chances after 32 . . . �xf2+ 33 �xf2 l:txd6 34 i.xd6 l:txe8.

33 �xb6 i.c4 34 i.b4 l:txe1+ 35 i.xe1 1-0

For a long time the schemes of development examined in this chapter were overshadowed by the system with l:tbl . As the popularity of that system wanes, the interest in the alternative variations can be expected to grow, and their intensive development promises the appearance of new discoveries and ideas.

5 Modern Exchange Variation with :b l

(Chapter introduction and Game 1 3 contributed by Viacheslav Eingorn.)

We now move on to one of the most topical systems within the Griinfeld, and arguably its main line for much of the past couple of decades. This line is characterized by White playing the Exchange Variation with the moves lLlf3 and l:tb1 . The first of these moves is natural enough, but the latter requires some explanation. Initially the move l:tbl has a rather outlandish appearance, although the fact that the move is such a well-established part of the theoretical landscape may blind us to just how surprising the move seemed in its early years. While the move l:tc 1 (the main approach at the start of the 1 980s) has a clear purpose, putting the rook on b l fails to defend c3 and targets a defended pawn. Yet it has become by far the most popular method, and has proved an enduring weapon, giving Griinfeld players serious headaches for many years, with no truly clear-cut solution found despite extensive testing at the highest levels.

In fact, the .l:lbl idea arose from a consideration of the drawbacks to other moves. The birth of this system was preceded by a crisis for White in the variation I d4 ltJf6 2 c4 g6 3 lL!c3 d5 4 cxd5 ltJxd5 5 e4 ltJxc3 6 bxc3 i.g7 7 ltJf3 c5 8 i.e2. This bishop move looks extremely natural, but after 8 . . . ltJc6, attacking the pawn-centre without delay, White lacks an effective reply. 9 i.e3 is well met by 9 . . . i.g4! , when White's centre becomes a static target rather than a dynamic weapon, while 9 d5 i.xc3+ 1 0 i.d2 i.xal I I 'ii'xa1 ltJd4 12 lLlxd4 cxd4 1 3 'ii'xd4 is unconvincing after either 1 3 . . . 0-0 or I 3 .. .f6. The check 8 i.b5+ also fails to pose Black significant problems because of the same reply, 8 . . . lL!c6; e.g., 9 0-0 cxd4 I O cxd4 0-0 I I i.e3 i.g4.

With those lines in mind, the apparently unpretentious 8 l:tb I turns out to be a rather nimble way to parry Black's main active ideas. Now . . . ltJc6 can be met by d5, while . . . i.g4 leaves the b7-pawn en prise. White intends to play i.e2 next, and it is hard for Black to find a reply that is quite as useful as .l:tbi that serves as preparation for . . . ltJc6 or . . . i.g4. Normally Black simply castles, and 8 . . . 0-0 9 i.e2 (D) brings us to the standard position.

B

The result of the insertion of the moves .l:tbl and . . . 0-0 is that White's centre retains its dynamic mobility. There are many options for Black here, with no clear consensus as to which is best: • 9 . . . b6. Black avoids an early confrontation, preferring to continue the development of his pieces

for the time being.


• 9 . . . lbc6 is, as advertised, met by 10 d5. As White has removed his rook from the g7-bishop's line of fire, this continuation now involves the sacrifice of the c3-pawn only. Although 10 . . . ..txc3+ is playable, the reserved 10 . . . lbe5 is more sensible.

• 9 . . . ..tg4. This is the other natural method to increase the pressure on the enemy centre, but after 10 0-0 cxd4 1 1 cxd4 ..txf3 12 ..txf3, White's chances tum out to be preferable.

• 9 .. :ii'a5 and 9 . . . cxd4 l 0 cxd4 11i'a5+ constitute Black's most radical reaction - he seeks to disturb the material balance in order to give the struggle a concrete tactical quality. All of the continuations listed above gradually went through scrupulous analytical and practical

tests, which resulted in the creation of the impressively sized section of the Grtinfeld Defence under the conventional title of 'l:.b 1 ' . Many of these lines have been worked out in great detail, with novelties being introduced well beyond move 20. As someone who stood at the very source of this whole variation three decades ago when it was still a vast expanse of virgin territory, I (V.E.) now watch its unrestrained evolution not only with interest but also with a certain degree of irony.

Game 13 (Eingom-Malaniuk) features 9 . . . b6, and serves as an introduction to some of the key ideas behind .l::tb1 .

In Game 14 (Vai·sser - Vachier-Lagrave), we move on to 9 . . . lbc6 10 d5 ..txc3+ with Black grabbing a pawn at the cost of allowing White a serious initiative, while also discussing Black's 8thmove alternatives to castling.

The subject of Game 15 (Shabalov-Areshchenko) is 9 . . . lbc6 10 d5 lbe5, a line that can lead to extremely forcing play, with plenty of spectacular ideas. These were among the earliest l:.b1 lines to be analysed in great depth, with one particularly notorious subvariation featuring an 'opening' novelty on move 36 back in 1984.

The final example of the chapter, Game 16 (Gelfand-Shirov), deals with 9 . . . 1!i'a5 and 9 . . . cxd4 10 cxd4 1!i'a5+. The latter has been the most popular line in the more recent history of .l::tb l , with the pawn sacrifice 1 1 ..td2 taking centre stage.

Game 1 3 [085]

Viacheslav Eingorn - Vladimir Malaniuk Baku 1 9 79

1 d4 lbf6 2 c4 g6 3 lbc3 d5 4 lDf3 ..tg7 5 cxd5 lbxd5 6 e4 lbxc3 7 bxc3 c5 8 .l::tb1 0-0 9 ..te2 b6 (D)

w

By refraining from the attack on the d4-pawn (9 . . . lbc6) and the diversionary . . . 11i'a5, Black wants to turn the game down a relatively quiet course. Still, we can point out that in comparison with the classic variation 4 cxd5 lbxd5 5 e4 lbxc3 6 bxc3 ..tg7 7 ..tc4 c5 8 lbe2 White's minor pieces occupy more favourable positions, which accentuates the importance of his pawn-centre.

10 0-0 The premature flank attack I 0 h4 is success

fully parried with 1 0 . . . ..tg4. Castling appears the optimal decision, although 10 ..te3 ..tb7 also deserves consideration and now either I I e5 with a subsequent h4, or I I 1!i'd3.

MODERN EXCHANGE VARIATION WITH :bJ 63

10 ••• ..tb7 Various experiments, like the line 10 . . . 1!i'c7

1 1 ..tg5 ..tb7 l 2 11i'd3 h6 ( l 2 . . . e6 ! ?) 13 ..th4 ( 1 3 ..te3 i s not bad either) l 3 . . . cxd4 14 cxd4 'ii'f4, are not appropriate in this situation, because in the end they are merely grist to White's mill: 1 5 d5 ..ta6 l 6 11i'd 1 11i'xe4 1 7 ..txa6 lt::lxa6 1 8 :et 'ii'f5 19 ..txe7 :fc8 20 d6, Haba-Feher, Zalakaros 2000.

ll 'ii'd3 (D) A noteworthy manoeuvre: the queen pro

tects the pawns on e4 and c3 while preparing the d5 advance. I I 'ii'c2 is less expedient because of l l . . .cxd4 12 cxd4 'ii'c8, but the immediate I I d5 ! ? (a thematic pawn sacrifice) is interesting: 1 l . . . ..txc3 1 2 ..tc4 with adequate positional compensation for the slight material loss.

B

1 1 ••• cxd4 This type of exchange conforms to the spirit

of the position in principle, but there was no need to hurry with it. In the current situation Black usually employs one of these two approaches: he either restrains the d4-pawn with . . . e6, or tries to deflect White from advancing with d5. Here are a few sample lines (in 'barebones' style - you are invited to study the lines further if they attract your interest) where Black for the moment refrains from the exchange of pawns:

a) l l . . .e6 1 2 ..tg5 ( 1 2 :d1 lt::ld7 1 3 ..tg5 11i'c7) and now Black can try 1 2 . . . 11i'c7 1 3 11i'e3 ( 1 3 d5 exd5 14 exd5 'ii'd6) 13 . . . lt::ld7, while an alternative set-up is 12 . . . 11i'd6 1 3 11i'e3 cxd4 (or l 3 . . . :c8) 14 cxd4 lt::lc6.

b) l l . . . ..ta6 1 2 1!i'e3 and now Black has played l 2 . . . 11i'c8, 12 . . . e6 and l 2 . . . ..txe2, but l 2 . . . 1!i'd7, intending to generate counterplay by

. . . 1!i'a4, is more interesting. White has several replies, including 13 dxc5, 13 d5 ( l 3 . . . ..txe2 14 'ii'xe2 ..txc3 15 ..th6) and 13 ..txa6 lt::lxa6 14 d5.

c) l l . . .'ii'd7 12 d5 !? ( 1 2 ..tg5 and 12 :d1 are also possible) 1 2 . . . c4 1 3 'ii'c2 :c8 14 lt::ld4 is another line.

Assessing of the pros and cons of the inclusion of the preliminary l l . . . cxd4 12 cxd4, it is hard to draw a definite conclusion. Although this way Black succeeds in restricting somewhat his opponent's choice, given an opportunity, White can tum the quick opening of lines (the c-file, the a3-f8 diagonal) to his advantage.

12 cxd4 e6 (D) l 2 . . . ..ta6 and l 2 . . . 1!i'd7 still remain the alter

natives.

w

13 ..tgS White takes advantage of the favourable op

portunity to regroup with ..tg5 and 'ii'e3. The continuation 1 3 l:td I is about as good.

13 ••. 11i'd6 Here the move to c7 or c8 is out of the ques

tion, but the queen is not safe on d6 either, as it must constantly watch for a new attack.

14 1!i'e3 Having received the necessary support, the

bishop, on the other hand, is placed very well on g5, controlling the important dark-squared diagonals. This permits White to plan the advance on the flank with h4-h5, in addition to operations in the centre, whereas Black still needs to complete the mobilization of his forces.

14 ••• lt::ld7?! A substantial positional error. l4 . . . lt::lc6 was

preferable, although in this case White keeps the initiative with 15 :fd l :fc8 ( l 5 . . . :fe8 16 ..tb5; 1 5 . . . :ac8 16 h4!?) 1 6 d5 (or 16 ..th6). Black could have also begun with 14 ... :c8, in


order to continue after 15 l:tfdl with 1 5 . . . tt::lc6 or even 1 5 .. ."ii'f8. The move chosen by Black is inferior because, above everything, the f6-square is not available to the knight thanks to the threat of a pawn fork, while if left on d7, the knight only interferes with the coordination of the rest of the pieces.

B

15 l:tfd1 .l:tfc8 16 �b5! (D) More accurate than 16 h4 Vf8.

Now the other white bishop takes up a menacing position too, and on top of that it cannot be driven away, because after 16 . . . a6 17 �xd7 'ii'xd7 the b6-pawn is hanging.

Somehow Black has imperceptibly drifted into a position where he experiences a severe shortage of constructive moves. For example, 1 6 . . . �c6 is unacceptable due to 17 �a6. Now the objectively best continuation is probably the modest 16 . . . lbf8, even if such a passive retreat constitutes a tacit acknowledgement of White's evident positional advantage.

16 .•. .l:tc2?! Curiously enough, this pseudo-active sortie

occurred again in several later games. In the vain search for counterplay, Black pushes too hard and, as a result, merely aggravates his problems. We should add that the other attempt to complicate the fight promises nothing good either: 16 . . . h6 17 �xd7 ( 17 e5 is not as clear) 1 7 . . . Vxd7 1 8 �xh6 �xh6 19 Vxh6 �xe4 20 tbe5 (20 lbg5 �xb1 21 Vh7+ 'it>f8 22 .l:txbl l:tc7 !) 20 ... Ve7 2 1 l:tb3, and White's attack on the black king then develops without a particular effort.

17 �a4 At once inviting the rook to clarify its inten

tions. Strictly speaking, it should have now returned whence it came.

17 ... l:txa2 18 e5 Vd5 19 �b3 VaS It is not hard to see that Black's exchange

sacrifice is not quite sound. Even the relatively simple 20 .i.xa2 Vxa2 2 1 lbd2 (2 1 l:tbc l ! ?) 2 1 . . . Vd5 22 f3 f6 23 exf6 lbxf6 24 �xf6 �xf6 25 lbe4 �g7 26 'it>h I favours White, but he had another idea in mind.

20 d5! (D)

B

The pawn breakthrough in the centre is quite a tempting idea, as Black's major pieces lined up along the a-file don't take part in the ensuing battle.

20 ... �xd5 21 i.xd5 exd5 22 e6 lbf6 After 22 . . . lbc5 23 exf7+ 'it>xf7 24 l:txd5

White's attack is decisive. The main line runs as follows: 24 . . . l:tal 25 Ve7+ 'it>g8 26 l:td8+ l:txd8 27 Vxd8+ 'it>f7 28 lbe5+! 'it>e6 29 Vg8+ 'it>f5 30 Vf7+ 'ittxg5 3 1 h4+, mating. The retreat 22 . . . lbf8 doesn't save Black either. There follows 23 exf7+ 'it>xf7 24 l:tdc l ! (24 l:tbc l ? ! allows resistance by 24 . . . 'it>g8 25 l:tc7 Vb5 ! 26 Ve7 Vb2, although 27 'it>h l ! 'it>h8 28 g3 l:tal 29 l:txal Vxa1 + 30 'it>g2 should be winning), when 24 . . . lbe6 loses to 25 tbe5+ �xe5 26 Vxe5 l:te8 27 l:tc8, and after 24 . . . 'it>g8 25 l:tc7, Black no longer has the 25 . . . Vb5 defence, while 25 . . . l:tal simply loses to 26 l:txg7+.

23 exf7+ 'it>xf7 24 l:tbc1 l:te8 25 l:tc7+ It would appear that the invasion by the white

rook should quickly conclude the struggle, but quite a few adventures still lie ahead.

25 ... 'it>f8 25 . . . 'it>g8 is no good because of 26 l:txg7+

'it>xg7 27 �xf6+. But the text-move should not have fared much better . . .

26 Vct? Stumbling over a flat spot. At least two lines

guarantee White victory: the complex 26 Vf4

MODERN EXCHANGE VARIATION WITH '!:..bl 65

lta1 27 'ili'd6+ 'itg8 28 ltxg7+ 'it>h8 29 g4! .l:.xd1 + 30 'it>g2 and the simple 26 �xf6 l:txe3 27 �xg7+ Wg8 28 fxe3.

26 .•. lL!h5! (D)

w

The only reply that allows Black to prolong the resistance. White's desire to reach the goal by the simplest means has played a cruel trick on him and noticeably complicated his task.

27 �h6? Yet another oversight. It seems that the win

was still within reach after 27 'i!i'c6! (27 �e7+ looks somewhat weaker) 27 .. . 'ili'c5 (27 .. J:tae2 28 g4) 28 'i!i'xe8+ (or 28 ..ie7+ 'i!i'xe7 29 ltxe7 .l:.xe7 30 'ili'a8+ l:le8 3 1 'i!i'xd5) 28 . . . 'it>xe8 29 .l:.e1 + 'itf8 30 �e7+.

27 ... 'i!i'a4 Serious problems await Black in case of

27 . . . l:.ae2?! 28 g4 'i!i'a4 29 .l:.d4. Even though the text-move gives him a satisfactory defence, 27 . . . 'i!i'b4! is more accurate, bearing in mind the line 28 �xg7+ �xg7 29 'ili'h6 'ili'b2 30 l:lfl .l:.a1 . And if White replies as in the game, with 28 l:.fl , then the black queen is posted better on b4.

28 .:.n .l:.ae2 29 ..txg7+ lLixg7 30 'i!i'gS By playing 30 'i!i'c3 d4 3 1 l2Jxd4 .l:.2e7 32

lL!e6+ 'it>g8 33 lLixg7 .l:.xc7 34 'i!i'xc7 White retains an extra knight for a while, but it is not clear how he can extract any real benefit from this circumstance. He decided to maintain the tension, but 30 'ili'h6!? .l:.2e7 3 1 .l:.xe7 ltxe7 32 'i!i'xh7, devoting maximum attention to the enemy king, was perhaps more suitable for the same purpose. The subsequent events took place in severe time-trouble and hardly need a detailed commentary.

B

30 ... .l:.2e6 31 'ii'xd5 (D)

31 .. .'i6'e4?! Equality results from 3 l .. . lt6e7 32 lLig5 lLif5. 32 'ii'a2 h6?" Here, too, 32 ... .l:.6e7 is better. Considering

the likely loss of both queenside pawns, Black should have avoided as much as possible creating additional weaknesses.

33 'i!i'xa7 .l:.a8? The final oversight, leading to an immediate

defeat. 34 .l:.f7+ 1-0

Game 1 4 [085]

Anatoli Va"isser - Maxi me Vachier-Lagrave French Ch, Chartres 2005

1 d4 lLif6 2 c4 g6 3 lLic3 d5 4 cxd5 lLixdS 5 e4 lLixc3 6 bxc3 �g7 7 lLif3 c5 8 ltb1 0-0

There is little point in 8 . . . 'ii'a5 before castling. A possible, though not mandatory, reply is 9 l:tb5, when 9 . . . 'ii'xc3+? is bad due to 10 ..id2 'i!i'a3 1 1 'ii'c2, when there is no apparent defence against J:tb3. Better is 9 . . . 'ii'xa2 10

.l:.xc5 lL!d7 with an unclear position; White controls the centre, but the c3-pawn is backward, and the rook is out on an extended trip. 9 ..id2 is played more often, generally transposing into the main variations (see Game 1 6) .

The same can be said about 8 . . . cxd4 9 cxd4 'ii'a5+ 10 �d2 'ii'xa2. White can return to the


well-trodden path of Game 16 if he wants, but now he has the extra option of developing the bishop more aggressively by 1 1 'ii'c 1 0-0 12 ..tc4.

We shall examine 8 . . . tt:lc6 closely. White's thematic and critical reply is of course 9 d5 . Then yet another type of endgame is possible: 9 . . . ..txc3+ 10 ..td2 ..txd2+ 1 1 'ii'xd2 tt:ld4 1 2 tt:lxd4 cxd4 1 3 'ii'xd4 (the advantage remains with White in the major-piece middlegame as well - 1 3 ..tb5+ ..td7 14 ..txd7+ 'ii'xd7 1 5 'ii'xd4; the choice i s a matter of taste) 1 3 . . . 'ii'a5+ 14 'ii'd2 'ii'xd2+ 1 5 �xd2 (D).

B

We shall encounter a similar ending later; the difference here is that the black king is still in the centre. Does this provide more grounds to hope for equality? From d8 the king covers the point of entry but disconnects the rooks, and he won't be allowed to settle on d6. Therefore, here too, White's remaining pieces will be much more active than those of his opponent.

Instead of the capture on c3 Black has another standard option - 9 . . . tt:le5, but here in addition to transposition into the main continuations that we examine in the current game, White has the additional possibility of a check on b5 with the subsequent formation of a light-squared pawnwedge in the centre, but now without the passive light-squared bishop.

9 ..te2 (D) 9 •.• tt:lc6 The other method of putting pressure on the

centre - 9 . . . ..tg4 - has not received general acceptance. As a rule, White agrees to swap the d4-pawn for the one on b7, but without allowing Black to centralize his knight. Then the bishoppair advantage secures White a small but solid initiative: 1 0 0-0 ..txf3 (here, as well as after a

B

preliminary pawn exchange, the knight move to c6 provokes the d5 advance) 1 1 ..txf3 cxd4 1 2 cxd4. In response to any capture on d4, White can take the b7-pawn, while 1 2 . . . ..txd4 may be met with the sharper 1 3 e5 tt:lc6 14 ..th6 lieS 1 5 e6.

10 d5 ..txc3+ 1l ..td2 ..txd2+ 12 'ii'xd2 tt:la5 The endgame that arises after the return of

the pawn by 1 2 . . . tt:ld4 1 3 tt:lxd4 cxd4 14 'ii'xd4 'ii'a5+ 15 'ii'd2 'ii'xd2+ 16 �xd2 (D) was rejected by theory long ago.

B

It is easy to see that the only differences between this diagram and the one at the beginning of the game are that Black has castled and White has the extra move ..te2. The black king is farther away from the centre, and in particular from the defence of c7; on the other hand, it allows the rook into the battle. The e2-square is not the most functional post for the bishop, but this inclusion is undoubtedly useful if only because the rook's path has been cleared.

Why is White, so reluctant to exchange the queens in other cases, so happy to go for the endgame in this edition? The key point is that the versions of the exchange that we have already seen did not lead to an endgame; those

MODERN EXCHANGE VARIATION WITH 'f:.bl 67

positions could be classified as a queenless middlegame as there were many pieces left on the board, and the white king was uncomfortable in the centre. However, here it is ready to take a most active part in the fight and work-wise it is quite a powerful unit. The other factor that tips the scales in White's favour is his advantage in development, and the fewer pieces remain, the more it makes itself felt. As an example, let's have a look at the source game of this ending, Eingorn - Sideif-Zade, Tashkent 1 980: 16 . . . l:td8 (the immediate 16 . . . e6 is well met with either 17 .i.c4 or 1 7 d6) 17 'i&?e3 b6 1 8 l:tbc I (precisely this rook, it has nothing left to do on b1) 18 . . . e6 1 9 .i.c4 'i&?f8 20 l:thd I (20 dxe6 promises an advantage as well, but White strives for the maximum) 20 . . . 'i&?e7 21 e5 (a refined move, fixing dark squares and cutting off the enemy's retreat; because of the threat of d6+, the reply is forced) 2 l . . .exd5 22 .i.xd5 l:tb8 23 :tc7+ .l:.d7 24 l:tdc I .i.b7 25 .i.xf7 .i.xg2 26 .l:.xd7+ 'it>xd7 27 'i&?f4 and the decisive penetration by the king to f6 cannot be prevented.

We now return to 12 . . . ttJa5 (D):

w

So then, White has given up a pawn. As compensation he has the superiority in the centre, the enemy king is deprived of its main defender, and the aS-knight is on the rim. The initial impression is that this is more than enough and the position favours White, which was the general consensus in the early days of the variation's history. As the role of computers in analysis grew, so did the confidence in defensive resources; the machines themselves, with their less pious perception of classic chess laws, grab pawns in much worse situations. As a result, a certain reassessment of the diagram position took place, and today the variation is

recognized as playable and is in a state of active development.

13 h4 This move suggests itself - the threat is a

conventional mating attack. 13 ... .i.g4 (D) It might seem that the attack can be re

pulsed with 1 3 .. .f6, covering the dark squares, opening a path for the rook and the queen, intending to meet 14 h5 with 14 . . . g5. However, White then has 15 ttJxgS, when accepting the sacrifice loses: 15 . . .fxg5? 1 6 'i!kxg5+ �h8 (or 1 6 . . . �f7 1 7 h6 'i&?e8 1 8 .i.b5+ .i.d7 1 9 'i!kh5+ .l:.f7 20 e5) 17 h6 l:tf7 1 8 .i.h5 'i!kg8 1 9 'i!kxg8+ (also good is 19 'ii'e5+ l:tf6 and, for example, 20 g4) 1 9 . . . �xg8 20 .i.xf7+ 'i&?xf7 2 1 l:tc l b6 22 l:tc3, Savchenko-Svidler, Groningen open 1 993. The combined attack by the two rooks allows a quick win of the h7 -pawn, after which there are only two steps separating the squaddie on h6 from the general 's epaulettes. Black can hold out by declining the offer and playing 15 . . . e5, but then the extra pawn is gone, while White retains chances of developing his initiative.

w

The top priority is the elimination of the f3-knight, the most dangerous assault unit.

14 ttJg5 This move keeps the knight, but the price is

the speed of the offensive. Let's examine the other option, 14 hS. It is obvious that 14 .. . .i.xh5? is no good, as both 15 'i!kh6 and 15 g4 win easily. Therefore Black must pick the most radical method of forestalling the potential threat of ttJgS, namely 14 . . . .i.xf3 1 5 gxf3 (it is necessary to keep the c4-square under control, since after 1 5 .i.xf3? ttJc4 16 'i!kh6 'ii'a5+ Black launches a counterattack) 15 . . . e5. In order to protect the


h7-pawn, Black needs to clear his second rank. The idea of the defensive set -up is based on the dark-squared blockade of the central pawngroup, so that the pawn-chain should remain on the light squares, curbing the e2-bishop' s activity. The knight is planned to be transferred to d6. If Black manages it all in good time, he will have sufficient counterplay, for the sake of which the extra pawn can be returned as well; for example, 16 'ifc3 l:te8 17 hxg6 fxg6 1 8 'ifxc5 b6 1 9 'ife3 l:tc8. For now the attempt to lift the blockade immediately by 1 6 f4 fails to bring success after 1 6 . . . exf4 17 'ifxf4 'ife7, so the search for an advantage is focused on 1 6 hxg6 fxg6 1 7 d6 (D).

B

Here, too, White wins back the pawn, because 1 7 . . . l:te8? is bad due to 18 'ifd5+ �g7 (or 1 8 . . . �h8) 1 9 .:txb7+, while 17 . . . l:tf7? blunders the knight away: 1 8 'ifd5 'iff6 1 9 l:tb5. The main continuation is 1 7 . . . b6 18 'ifd5+ �g7 19 'ifxe5+ 'iff6. If the queens are exchanged, then White's own king in the centre doesn't let the white pieces provide the pawns with effective support, and it is not clear even how to save them. Thus the logical course is to combine the threat of the pawn advance with play against the enemy king. However, Black, whose knight has been given a direct route to the unassailable d4-square, also completes the centralization of his army and for the moment successfully fends off White's offensive: 20 'ifh2 h5 21 l:.dl tiJc6 22 .:tgl tiJd4 23 l:td3?! tiJe6 with a new blockading scheme, Zhou Jianchao - Vachier-Lagrave, China-France match, Chenzheng 2005.

14 ... .i.xe2 15 'it>xe2 h6 The other continuation is 15 . . . e6 1 6 h5 tiJc4

17 'ifcl exd5. White has two branches to explore:

a) 1 8 tiJxh7 .:te8 1 9 hxg6 l:xe4+ 20 'it>fl l:h4 21 .l:.xh4 'ifxh4 22 �g1 tiJe5 23 g3, ErnstL' Ami, Groningen 2004.

b) 1 8 hxg6 fxg6 1 9 l:txh7 'iff6 20 f4 b6 21 .:tb3 l:lae8 22 l:tbh3 l:te7 23 l:t7h6 and it is not easy for Black to disentangle himself, AvrukhSutovsky, Ashdod 2004.

16 ti)f3 This retreat slows down the tempo of the at

tack. 16 h5 hxg5 1 7 hxg6 fxg6 has also been tested. The capture 1 8 'ifxg5 takes the heat off the a5-knight and allows the black queen to start harassing the white king by 1 8 . . . 'ifd6 19 l:.h6 'ifa6+ 20 �e1 'it>g7. Beginning with 1 8 l:.h6(?) at first looks more cunning, as doubling the rooks on the h-file is threatened too. However, if Black finds 1 8 . . . l:r.xf2+ 1 9 'it>xf2 'iff8+ 20 �g1 'ifxh6 21 'ifxa5 g4, it is the white king who is in trouble. The game continuation is also possible after the retreat 1 6 tiJh3.

16 ••• �h7 17 'ifc3 b6 18 tiJgS+ (D)

B

The knight offers itself on the same square, although in a different edition. Taking it at once is no good, but it doesn't plan to retreat a second time.

18 ••• 'it>g8 19 hS hxgS The subsequent events give Black no cause

for optimism, so we shall mark the recently employed 1 9 . . . 'ifc8 as deserving consideration.

20 hxg6 fxg6 21 l:h8+ It would be interesting to test 2 1 .:tbd 1 , threat

ening to advance the centre pawns. 21...�f7 22 l:th7+ �e8 23 'ifg7 �d7 23 . . . 'ifd7?! blocks the king's natural retreat

path and endangers the f8-rook; 23 . . . 'ifd6?!, inviting 24 e5, is still more dubious. The textmove gives White a strong passed pawn.

24 d6 'ife8?!


Since this turns out badly, one can suggest 24 .. J::te8 ! ? 25 dxe7 'fkc7.

25 dxe7 .l::tg8 26 We5 The e7-pawn creates tactical tension as with

the black king on its second rank, there are ideas of queening with discovered check. The pawn has to be blockaded, which gives rise to combinative motifs to clear its path.

26 ••• <it>c8 (D) Believe it or not, this move was an attempt

to improve over a previous game. The natural 26 . . . l2Jc6?! had been played, but it turns out to be a losing move: 27 .l::td I + l2Jd4+ 28 l:xd4+ cxd4 29 'fkd5+ 1;c7 and the left-hand rook was dealt a thematic blow - 30 'fkxa8.

w

27 'fkd5! 27 l:dl is possible. Black has to defend

against the check on d8 by 27 . . . <it>b7 28 l:d8 Wb5+ 29 1;f3 l:te8 30 'fkd5+, when a sharp major-piece endgame can arise: 30 . . . l2Jc6 3 1 l:xa8 1;xa8 32 1;g3 gives White at least a safe draw, while 30 . . . Wc6 is likely to lead to a repetition of moves in the line 3 1 l:f7 'fkxd5 32 exd5 l2Jc4 33

.l::tf8 l2Jd6 34 :d7+ ..ta6 35 l:d8 ..tb7. With the text-move, White hopes for more.

27 ••• l2Jc6 28 .l::tbh1 It seems that White might be angling for a

repetition by 29 .l:!.h8 rj;c7 30 .l:!.8h7 (threatening to take on g8) 30 . . . ..tc8. However, the position contains deeper ideas than that.

28 .•• l2Jd4+ 29 'it>e3 l:tb8 30 g4 (D)

B

It turns out that Black is in zugzwang: the king cannot venture onto its second rank because of 'ii'xg8, and the queen has to defend the g8-rook, which has not a single move left. Why the other rook cannot go to b7, we shall find out from the game. The knight must defend the e6-square and is only capable of driving the white king to g2, and then what? The a-pawn can only manage a few steps.

30 ••• .l::tb7 31 'ii'xg8 This time the right-hand rook receives the

symmetric blow. The rest is simple. 31 ••• 'fkxg8 32 .l:!.h8 .l::txe7 33 1hg8+ ..tb7 34

.l::txg6 .l::tf7 35 .l::th3 l2Jc2+ 36 <it>e2 l2Jd4+ 37 <it>n 1-0

Game 1 5 [0851

Alexander Shabalov - Alexander Areshchenko Port Erin 2005

1 d4 l2Jf6 2 c4 g6 3 l2Jc3 d5 4 cxd5 l2Jxd5 5 e4 l2Jxc3 6 bxc3 1J..g7 7 l2Jf3 c5 8 .l::tb1 0-0 9 1J..e2 l2Jc6 10 d5 l2Je5 (D)

U l2Jxe5 White's centre has been forced to advance,

temporarily ceding control of the e5-square. White plans to regain control of this square, and thus restore his centre's mobility, by a later f4

advance. Thus the exchange of knights, clearing the f-pawn' s path and luring the bishop out in order to gain a tempo on it, looks absolutely logical. There is some sense in avoiding the knight exchange, in order to keep more pieces on the board; then Black, who has less space, will be more restricted in manoeuvring. But there is no convenient retreat, since 1 1 l2Jd2


w

exacerbates White's lag in development, and Black replies 1 l . . .f5 and retains active chances.

ll ... �xe5 12 'i!i'd2 While not entirely natural, this is nevertheless

the main continuation. For instance, in reply to 1 2 �d2 Black plays 1 2 . . . 'i!i'c7, taking control of the h2-b8 diagonal, hampering both White's castling and f4; Black can then undermine the centre with . . . e6 and use the (half-)open e-file for counterplay.

1 2 c4 doesn't enjoy a great deal of popularity. This is not because it involves a pawn sacrifice ( 1 2 . . . 'i!i'a5+ 1 3 �d2 'i!i'xa2 takes a lot of time, and White will obtain compensation), but because the advance of the c-pawn prematurely cedes the d4-square to the enemy bishop, which can be reinforced with . . . e5. Despite such a formal minus as the pawns fixed on the squares of the same colour, it is hard to call the d4-bishop bad, unlike the e2-bishop which suffers from the same affliction. The move chosen by White protects c3 and at the same time makes the immediate advance of the fpawn inevitable.

Sometimes 12 l:tb3 is played, with the idea of transferring the rook to the kingside. This possibility should be paid due attention; as an example we shall mention the game S.lvanovV.Mikhalevski, St Petersburg-Beersheba match 1 999: 1 2 . . . e6 1 3 f4 �g7 14 c4 l:te8 ( 14 . . . �d4 is met with 15 �b2) 1 5 e5 f6 16 f5 (D) and Black failed to repulse the attack.

We shall see a similar method employing a battering-ram later.

12 .•• e6 Nowadays the pawn sacrifice 12 . . . c4 is un

dergoing practical tests. If the sacrifice is accepted, the extra pawn is hard to keep: 1 3 �xc4 'i!i'c7 14 'i!i'e2 a6 and now 15 �d2 seeks to

B

contain Black, but allows him the initiative by 1 5 . . . �d7 intending . . . �a4. In Khalifman-Belov, World Cup, Khanty-Mansiisk 2007 White returned the material by 1 5 0-0 �xh2+, but the weakening of the king's cover offset the superiority in the centre. It is wiser to fight for an advantage with 1 3 f4 or 1 3 0-0.

1 2 . . . 'ili'a5 does not enjoy recognition; after the forced 1 3 l:tb3 Black does not succeed in taking advantage of the rook's position for effective counterplay.

The prophylactic retreat 12 . . . �g7 (D) is more interesting.

w

With regard to determining the central pawnconfiguration, the move 1 3 0-0 is considered to be a loss of time, and White strives to save a tempo on castling in the main line. The critical continuation is 13 f4, when, if it isn't to come down to a transposition of moves, Black replies with 13 . . . e5, and only then follows 14 0-0, not wasting time on the defence of the c3-pawn. After 14 . . . exf4 1 5 'ili'xf4 'i!i'e7 1 6 �c4, with the idea of opening the a2-g8 diagonal for the bishop, White's prospects look better. He also keeps the initiative in the line 1 3 c4 b6 ( 1 3 . . . �d4 is possible here as well, but after the preceding retreat,


such a move is a loss of time and is psychologically unpleasant to make) 14 .tb2 e5 1 5 0-0 'ii'd6 16 a4 f5 1 7 a5 fxe4 18 .tc3, GelfandSutovsky, Ashdod 2004.

The main rival to the text continuation is 12 . . . b6, removing one of the pawns from attack and overprotecting the other. After 1 3 f4 .tg7 (D), White has several set-ups at his disposal:

w

a) White bolsters the centre with 14 c4, but loses the possibility of activating the lightsquared bishop via this square. Black retains even chances by setting up a blockade on the dark squares after the exchange of the queens and the dark-squared bishops: 14 . . . e5 (here this move does not close the position; on the contrary, it clears the blockading squares) 15 0-0 exf4 1 6 'fi'xf4 'ii'e7 17 .tb2 .txb2 1 8 .:!.xb2 .td7 1 9 .td3 and now 1 9 . . . f6 or 1 9 . . . .:!.ae8 preparing . . . 'fi'e5.

b) 14 0-0 is naturally met by the freeing 14 . . . e6 (D). Now:

w

b1) The d-pawn can go forward by 1 5 d6, but supporting it is complicated, and the counterattack against the vanguard cut off from the main forces gives Black fully adequate chances

after playing 1 5 . . . e5 1 6 .tc4 .te6 at once or following the preliminary 1 5 . . . .tb7 (not fearing 1 6 e5 f6 1 7 'fi'e3 fxe5 1 8 fxe5 'ii'h4) 1 6 .tf3 e5.

b2) White can exchange on e6 with a subse-quent f5, viz. 1 5 dxe6, meeting 1 5 . . . .txe6 with 1 6 f5. Then taking on a2 is bad: 16 . . . 'fi'xd2? 17 .txd2 gxf5 1 8 exf5 .txa2 1 9 .:!.b2 ..td5 20 f6 .th8 21 .th6 followed by .:!.f4 with an extremely strong attack. Therefore a retreat is forced. If 1 6 . . . .td7, then 17 f6 ..txf6 1 8 l::txf6 'fi'xf6 19 'fi'xd7 and the bishop-pair plus darksquared holes in Black's position promise White a clear advantage. That leaves 1 6 . . . .tc8, but the ending after 1 7 'fi' xd8 .:!.xd8 1 8 .tg5 f6 1 9 .tc4+ �h8 20 .th4 .tb7 2 1 .:!.bel .:!.d2 22 .:!.f2 also fails to inspire Black with optimism. Having failed to find equality in these lines, Black began playing 15 . . . fxe6 1 6 'ii'e3 'ii'e7 17 .tc4 �h8. To prevent Black from getting rid of his weakness with ... e5, White himself plays 18 e5, and in the resulting position both sides have defects in their pawn-structure, with mutual chances.

b3) 15 .tc4. Black cannot tolerate the enemy bishop on d5, while yet another version of the 1 5 . . . e5 advance also favours White. However, a temporary pawn sacrifice provides sufficient counterplay: 15 . . . .tb7 1 6 dxe6 .txe4 1 7 exf7+ �h8 1 8 .:b2 'ii'xd2 1 9 .txd2 .:!.ad8 20 l:tel and now 20 ... .td3 2 1 .txd3 (or 2 1 .te6 b5) 2 1 . . . .:!.xd3 leads to a drawn rook endgame. Instead, 20 .. J:f.xd2? runs into a pretty refutation: 21 :xd2 .txc3 22 .:!.d8 ! and White gains a decisive tempo, Wells-Krasenkow, Politiken Cup, Copenhagen 1 996.

13 f4 .tc7 In the recent past Black's main retreat was

1 3 . . . .tg7 (D).

w


By refraining from . . . b6, Black saves a tempo that can be used in the counterattack against White's pawn-centre. The lines 14 dxe6 .i.xe6 and 14 .i.c4 �a5 now look more acceptable to him, while in case of 14 d6 he can return to line 'b1 ' of the previous note, where he has decent counterplay, by playing 14 .. . b6 15 0-0.

That's why it is only natural that 14 c4 (D) became the main continuation.

B

White's intentions are clear. If allowed to, he will castle, offer an exchange of the darksquared bishops (the order of these steps can be reversed), and then play e5 . In case of the exchange on d5 he must be ready to recapture with the c-pawn, and the other bishop will work well from f3.

In the lines we examined above, Black's principal counter-argument to this was the ... e5 advance. It is obvious that with . . . e6 already played, this move would now cost Black a tempo, with the result that he wouldn't have the time to set up a dark-squared blockade. His only try is to prove that the white army is not ready to support the centre in an effective manner, and that the centre can be destroyed. We shall examine the following continuations:

a) 14 . . . b6 prepares the development of the bishop to a6, from where it forces White to keep defending the c4-pawn, or to b7, putting pressure on the d5-pawn. Now:

a l ) 1 5 0-0 allows 1 5 . . . exd5 16 cxd5 .i.d4+ 17 �hi and now Black can play l 7 . . . .i.b7 with the threat of . . . f5, while the immediate l 7 . . . f5 1 8 e5 �xd5 1 9 .i.f3 �c4 20 .i.b2 is also possible, with unclear complications.

a2) 1 5 .i.b2 .i.xb2 1 6 .l:r.xb2 ( 1 6 �xb2 weakens the d5-pawn, and Black has an interesting sacrifice of a whole rook: 16 . . . exd5 1 7 cxd5

.l:r.e8 1 8 e5 �xd5 19 .i.f3 �c4 20 .i.xa8 �xf4 2 1 .i.f3 .l:r.xe5+ 22 �fl �e3 with ample compensation, Khuzman-Sutovsky, Tel Aviv 1 999) 16 . . . .i.a6 17 0-0 exd5 18 cxd5 .i.xe2 19 �xe2 and now 1 9 .. .f5 destroys the centre. After 20 .l:r.d2 fxe4 the topic of the discussion changes: now White has to prove the merits of the strong, yet isolated, passed d-pawn.

b) Black can exchange on d5 irrunediately by 14 . . . exd5 1 5 cxd5 .i.d4. The idea of the variation is based on this move - White's castling is delayed, and the frontal defence of the d5-pawn is disrupted. Nothing better than 1 6 .i.b2 comes to mind, and the search for counterplay follows in two directions:

b1) 16 . . . �h4+ (in order to open the diagonal, so that the possible capture of the d5-pawn may gain a tempo in comparison with playing 1 6 . . . �e7) 1 7 g3 �e7 1 8 e5 (unclear complications arise after 18 .i.f3 .i.f5) 18 ... �d8 1 9 .i.f3 .i.f5 20 .l:r.c l . Black's sorties are over; the achievements are the development of the bishop to f5 and forcing g3, a move that exposes the king's possible residence still more. The preparations for a further advance of the centre pawns require time, and as a counterbalance Black tries to rush his queenside pawns forward after 20 . . . .i.xb2 21 �xb2 �b6. Nevertheless, White's chances are to be preferred.

b2) The pin 16 . . . �6 also does not equalize. The ending after 17 .i.d3 c4 1 8 .i.xc4 .l:.e8 1 9 �xd4 �xd4 20 .i.xd4 .l:r.xe4+ 21 �d2 .l:.xd4+ 22 �c3 .l:.xf4 is difficult for Black, despite his extra pawn, as the passed d-pawn is not easy to stop.

c) 14 .. . .l:.e8 prepares an exchange of pawns. Here the rook exerts masked pressure on the e4-pawn but gives up the control of the f-file, along which active operations are planned. White must advance by 1 5 e5, when Black achieves his idea with 1 5 ... f6. The centre cannot be maintained, so White endeavours to transform the lost pluses into newly acquired ones, specifically the strong passed pawn, the more active pieces and the insecure position of the black king. For that a pawn must be sacrificed: 16 d6 fxe5 17 .i.b2 exf4 ( 1 7 . . . e4, so as not to open the f-file, has also been tested, but equality failed to materialize) 1 8 .i.xg7 �xg7 1 9 0-0 (D).

For the moment Black is two pawns to the good, it's his move and the question arises -can't he keep both?

MODERN EXCHANGE VARIATION WITH 'IJ..bl 73

B

c l ) 1 9 . . . g5 makes too many holes in the king's cover. White must take advantage of it after 20 .l:tfdl 'ii'd7 21 'ii'c3+.

c2) 1 9 . . . e5 opens the h3-c8 diagonal, and Black suffers material losses: 20 d7 'ii'xd7 2 1 'ii'xd7+ i.xd7 22 .l:txb7 .l:te7 23 .l:tdl .l:td8 24 i.g4.

c3) For a while the position after 1 9 . . . .l:tf8 20 .l:txf4 .l:txf4 21 'ii'xf4 'ii'f6 was considered tolerable, until a long line, resulting in an endgame with a white bishop versus three black pawns, crystallized: 22 'ii'e4 l:.b8 23 l:.fl 'ii'd4+ 24 'ii'xd4+ cxd4 25 .l:tbl i.d7 26 i.f3 b6 27 c5 .l:tc8 28 c6 i.xc6 29 l:.c 1 i.d7 30 llxc8 i.xc8 31 i.c6 (D).

B

The ending was thought to be drawn until Novikov employed a novelty on the 36th move (which was a record in the pre-computer chess) and proved that White wins: 3 l . . .�f6 32 d7 i.xd7 33 i.xd7 e5 34 �f2 e4 35 i.c6 �e5 36 h4!, Novikov-Tukmakov, USSR Ch, Lvov 1984.

The line we have just seen featured a catastrophe due to the pin along the 7th rank. Human ingenuity being what it is, this problem gave birth to the idea of avoiding it by 13 ... i.h8. Now after the bishop exchange the king will wind up on the

eighth rank, and it will be possible to defend the f4-pawn with . . . e5. A few games turned out well for Black, until the discovery of the blow that effectively buried this idea: 14 c4 l:te8 15 e5 f6 16 f5 ! ! (D), McCambridge-Hjartarson, Grindavik 1984.

B

It transpired that the position of the bishop on h8 presents White with an important \empo, and he has a decisive attack after 16 . . . gxf5 17 .l:tb3, and also 16 . . . fxe5 17 fxg6 (but not 17 fxe6? 'ii'h4+ 1 8 g3 'ii'e4), as well as 16 . . . exd5 17 e6 (or 17 fxg6).

We now return to 1 3 . . . i.c7 (D):

w

As Black's problems after the retreat of the bishop along the long diagonal grew, so did the popularity of the text-move.

14 0-0 As 14 c4 is impossible, White must acquiesce

to the recapture with the e4-pawn in case of an exchange on d5, or else play 14 i.c4, although on this square the bishop facilitates Black's counterplay. The usual reply is 14 . . . a6 (Black doesn't have enough time to set up the wellknown blockade with 14 . . . e5 because White replies 15 'ii'f2 'ii'f6 16 0-0 exf4 17 i.xf4 i.xf4


1 8 'ii'xf4 'ii'xc3 1 9 e5 with the advantage). Now:

a) After the exchange 1 5 dxe6 'il'xd2+ 16 .i.xd2 fxe6, thanks to the attack on f4 Black manages to carry out . . . b5, and his pawn-structure is in no respect worse than White's.

b) In reply to 15 a4 there is a choice be-tween 1 5 . . . 'ii'e8 1 6 1i'a2 l:tb8 and the pawn sac-rifice 1 5 . . . b5 16 axb5 axb5 17 .i.xb5 exd5 ( 1 7 . . . .i.b7 !?) 1 8 .i.c6 l:ta6 with decent compensation.

14 ... exd5 15 exd5 .taS (D)

w

This is the next stage of the voyage begun with the retreat to c7. The bishop has helped to relieve the tension in the centre but found itself removed from the weakened squares around its king. Given that it is hard for it to return to the fianchetto, the bishop attempts to aggravate the poor coordination of the white pieces by hobbling and attacking the c3-pawn. It is not easy for White to rearrange his forces, but it must be mentioned that he may find that a small force is sufficient for a successful attack; for example, a pawn on f6 and the queen.

16 f5 This is the main continuation today. Black is

forced to capture this pawn with his bishop, so White immediately solves the problem of how to activate the bl -rook.

But let's examine some of the other continuations that have been tested:

a) The most radical method of fighting the aS-bishop is 1 6 l:.b5. White has a strong attack after 16 . . . a6 1 7 l:txa5 'il'xa5 1 8 f5, but Black has the solid 16 . . . b6 17 l:txa5 bxa5 1 8 c4 l:tb8, when in reply to 19 .i.b2 Black starts operations on the b-file, blocking the diagonal with . . . f6 first, if necessary. If White decides to treat

himself to the c5-pawn with 1 9 .i.a3, the counter-sacrifice 1 9 ... l:tb4 secures Black an excellent position.

b) 16 .i.a3 is at first sight strange. What is it going to do there after 16 . . . b6? The logic is as follows: White needs to rearrange his forces, and he'd like to let the rook pass through but is reluctant to put the bishop on b2, as that fails to hinder Black's plans (which include the advances . . . b5 and ... c4, securing the bishop a deluxe base on b6), so White temporarily moves it to a3 (from where it can, given the right opportunity, force the exchange with .i.b4), especially because . . . b6 somewhat loosens the light squares, and the other bishop can seek a better future on b5 or f3.

c) 1 6 :b3 initiates another regrouping plan. White relieves the queen from the defence of c3 and intends to retreat it, to restore the pawn's mobility and to open the third rank for the rook. Of interest is the exchange sacrifice 1 6 . . . b6 1 7 'ii'd 1 .i.d7 1 8 c4 i.a4 19 .i.b2, when the bishops promise support for the passed pawn, whereas for the moment the black bishop on a5 is operating at half capacity.

d) 16 d6 is the main rival to the move in the game. An interesting debate has developed in the line 1 6 . . . b6 17 .i.f3 .i.f5 (at the right moment, when the rook has no decent move) 1 8 .txa8 .i.xbl 1 9 .i.c6 'ii'f6 20 .i.b2. Accepting the pawn offer by 20 . . . .txa2? is bad, since after 2 1 c4! 'ii'e6 22 f5 the attack is hard to repulse, so the best defence is 20 . . . .tf5 2 1 :el l:td8 22 l:te8+ l:txe8 23 .txe8 a6 24 a4 'il'd4+ (D).

w

This last move is not mandatory, of course. The win of a piece with 25 cxd4 .i.xd2 26 d7 .i.xd7 27 .i.xd7 _leads to a draw after 27 . . . c4, but the sharp four-bishop endgame leaves White


with the better chances: 25 'ili'xd4 cxd4 26 d7 b5 27 cxd4, Khalifman-V.Mikhalevski, St Petersburg-Beersheba match 1 999.

16 •.• i..xf5 17 .l:.xb7 'ilff6 (D) After the events of this game, the queen's

longer route to the long diagonal, i.e. 17 . . . 'ili'd6 1 8 i..c4 'ili'e5, should be recognized as the safer one. White succumbed to a decisive attack after 1 9 .l:.e 1 ? 'ilff6 20 l:lb3 'ilfh4 21 .tn c4 22 .l:.b5 i..b6+ 23 �h1 i..c7 in Obodchuk-Sutovsky, Turin Olympiad 2006. In a recent encounter our silicon friend preferred 1 9 .l:.f3 .l:.ab8 20 .l:.xa7 i..b6 2 1 .l:.a4 i..c7 22 g3 l:lb1 , but the black rook's position on White's back rank proved adequate compensation for the pawn in Deep Fritz-Deep Junior, Elista 2007.

w

18 g4 This aggressive thrust is a naturally tempting

idea, though it used to be underestimated. Previously 1 8 i..b2 was more common.

18 •.• i..xc3 19 i..b2 (D)

B

19 •.• i..xb2 It turns out that after 1 9 . . . i..xd2 20 i..xf6 the

passed pawn, combined with the attack on f7, secures White a perceptible advantage. In order

to save the piece, Black is forced to display considerable ingenuity.

20 .l:.xb2 c4 21 i..xc4 The rook must control the b6-square, while

the queen monitors d4. Anyway, a more sensible way to prevent . . . c3 is not apparent.

21. .. 'ili'h4 22 .l:.f4 i..xg4 23 'i!fd4 h5 24 d6 (D)

B

The strong passed pawn, the attack on f7 and the great difference in piece activity suggest that the position can be evaluated as clearly favouring White. If only the king had better cover . . .

24 ••• .l:l.ac8 Black trains his sights on the terribly strong

c4-bishop and neutralizes 25 d7. 25 'ili'e5 This move, forestalling all checks and threat

ening 'ilfe7, looks very strong but allows Black to rearrange his pieces. Black also has good drawing chances in the ending after 25 .l:.xf7 .l:.xf7 26 i..xf7+ �h7 27 i..c4 'ili'g5 28 h4 'ilfc 1 + 29 .tn 'ili'c5.

25 •• Ji'd8 26 .l:.b7 ..trs 27 'ili'e7 (D)

B

27 ••• .l:.xc4 Black's desire to eliminate such a bishop is

understandable, but 27 ... �g7, defending against


the capture on f7 by removing the king from • check, is simpler. 28 l:.xc4 'ifa5 29 d7 'ii'd2 (D) B

w

30 d8'if The king's voyage to the queenside by 30

l:.c8 'ifd4+ 3 1 �fl .i.h3+ 32 <ot>el 'ifgl + is too romantic, but the endgame doesn't promise much because Black's position is extremely solid.

30 .•. 'ifxd8 31 'ifxd8 l:.xd8 32 l:.xa7 l:.d1+ 33 'ita'2 l:.d2+ 34 �g3 .ie6 35 l:.ca4 (D)

35 •.• <ot>g7?! More accurate is 35 . . . l:.d3+ and only after 36

�f4 to play 36 ... <ot>g7. Now after 37 h4 �f6 it is White who is more likely to have problems as the h-pawn is in danger, while in the line 38 l:.a3 l:.d4+ 39 �g3 l:.g4+ 40 �h3 Black at the very least has a perpetual check.

36 h4 l:.d3+ 37 �f2 l:.d2+ 38 �f3 l:.d3+ 39 �g2

Black has fair drawing chances, but White could have at least tested him by 39 �e4.

39 ... l:.d2+ lfz.llz

Game 1 6 [085]

Boris Gelfand - Alexei Shirov Odessa (rapid) 2007

1 d4 lt'lf6 2 c4 g6 3 lt'lc3 dS 4 cxdS lt'lxdS 5 e4 lt'lxc3 6 bxc3 .ig7 7 lt'lf3 c5 8 l:b1 0-0 9 .i.e2 cxd4

The same idea of the queen raid can also be carried out without exchanging the pawns, viz. 9 . . . 'ifa5 (D).

w

Let's examine White's options:

a) I 0 'ifd2 foregoes any sacrifices but agrees to the exchange of the queens, which we shall encounter later. Black does not have to hurry with the exchange, and can deploy his forces according to the standard pattern - . . . a6, . . . b5, . . . lt'lc6, . . . l:d8, . . . .ig4 with pressure on the d4-pawn, taking advantage of the fact that it is temporarily hobbled as long as its neighbour on c3 is insufficiently protected.

b) 10 .id2 invites a return to the main line. Black would be wise to accept the invitation and include the exchange on d4; when White plays d5 the presence of the c-pawns on the board is rather in his favour, as the prop on c4 will reinforce the central bastion, while Black's queenside counterplay develops more slowly.

c) 10 .l:.b5 (D) is the move that one's hand automatically stretches out to make, but unlike the variation examined in Game 14, where the queen sortie to a5 took place a move earlier, the

MODERN EXCHANGE VARIATION WITH �bl 77

situation is more favourable to Black, whose castling is more useful than the bishop's development to e2. The capture of either pawn is possible here:

B

c I ) The one closer to the centre looks tastier, especially since the loot is real: IO ... 'ihc3+ I I i..d2 'ii'a3. Now a queen hunt is pointless, the c8-bishop is protected, and 12 'ii'c2 can be met with the simple 12 . . . cxd4. In case of 12 .l:.xc5 the attack on the d4-pawn secures Black an excellent game: 12 . . . l2Jc6 13 llc4 i..g4 14 d5 i..xf3 1 5 gxf3 l2Jd4. More cunning i s 12 lla5 'ii'b2 1 3 .l:.xc5 and now Black usually takes the pawn with 1 3 . . . 'ii'xa2, although White's compensation is not bad. The less mercenary 1 3 . . . l2Ja6 (as played by Leko), aiming for rapid development, has also been tested.

c2) IO . . . 'ii'xa2 is also possible. This doesn't win a pawn, but the queen's time-expenditure is balanced by the rook's expedition. Each side has to worry about the return of his commando home, and there are chances for both sides.

d) The most natural move is I 0 0-0, again offering Black two pawns to choose from:

d l ) Following 'tradition', let's start with the capture IO . . . 'ii'xc3. Now a position similar to the one examined in line 'c I ' can be reached via 1 1 l:.b3 'ii'a5 12 .l:.b5 'ii'xa2 1 3 .l:.xc5; the difference is in Black's favour, but this has no crucial importance. The main continuations for White feature either play against the enemy queen or play in the centre. In the first case the pawn has to be returned but the problems remain: 1 1 i..d2 'ii'a3 12 'ii'c2 i..d7 1 3 dxc5 i..c6 14 i..b5 l2Ja6 1 5 l:!fc 1 . The line I I d5 gives fewer causes for worry; for example, l l . . . 'ii'a5 1 2 i..g5 'ii'c7 1 3 'ii'd2 i..g4 14 e5 i..xf3 1 5 i..xf3 i..xe5 16 d6 i..xd6 1 7 l:txb7 'ii'c8 1 8 l:tbbl ? ! l2Jc6 1 9 'ifd5

l2Jd4?! ( 19 . . . .l:.b8!) 20 'ii'xa8 'ii'f5 2 1 'it'd5 l2Jxf3+ 22 'it'xf3 'ii'xg5, P.H.Nielsen-Carlsen, Wijk aan Zee 2005.

d2) This time the capture of the rook's pawn is more common: IO . . . 'ii'xa2. Now I I i..g5 gets the bishop to g5 in one move, without the stopover on d2, and it is harder for Black to equalize than with the main move-order. Now:

d2 1 ) The endgame resulting from I I . . . 'ii'e6 12 e5 ( 1 2 'it'd3 also promises an advantage) 1 2 . . . .l:.d8 1 3 'ii'a4 'ii'c6 14 'ii'b3 i..e6 1 5 c4 cxd4 1 6 i..xe7 llc8 1 7 l2Jxd4 i..xc4 1 8 l2Jxc6 i..xb3 19 l2Jxb8 i..e6 20 i..d6 is closer to a draw, Black can give up the exchange in two ways, but in both cases some suffering is unavoidable: 20 . . . .l:.axb8 or 20 . . . i..f8 2 1 i..f3 i..xd6 22 i..xb7 i..xb8 23 i..xa8 ..i.xe5.

d22) l l . . .l2Jd7 deserves more attention than it has received. After 12 i..xe7 l:te8 1 3 .i.d6 cxd4 14 l2Jxd4 l2Jf6 15 f3 the pawn-centre has disappeared, leaving White with the more active pieces, but Black is not without chances either.

10 cxd4 'ii'a5+ 11 i..d2 White gives up the a2-pawn but hopes to

show that the time gained, and the exposed position of the black queen, provide good compensation.

If White is not in a gambit mood, he must agree to the previously-mentioned exchange of the queens by 1 1 'ii'd2 'ii'xd2+ 12 ..i.xd2 (D).

B

The search for an advantage for White was carried out at the highest level, but Black held out; the line is considered acceptable for him and is seen rarely in the practice of recent years. To some degree, the explanation lies in the fact that for the subsequent struggle the move llbl turns out to be a loss of a tempo. With the queens gone, White has to forget

78 CHESS EXPLAINED: THE GRVNFELD

about attacking the enemy king, but there are still a lot of pieces on the board, and the attempt to obtain an 'extra king' in the centre is risky for White, so his majesty has to take a trip to the suburbs. From b 1 , an effective invasion on c7 cannot be arranged, so Black is able to neutralize the c-file. The pawn-centre will have to be transformed into a passed d-pawn, and it is this pawn that will be the main subject of the discussion. We shall illustrate this with the following example: 12 . . . e6 1 3 0-0 b6 14 .l:tfd1 .i.b7 1 5 d5 exd5 1 6 exd5 lt:'ld7 1 7 .i.b4 .l:tfc8 1 8 .i.e7 .i.f6 1 9 d6 �g7, Karpov-Kasparov, World Ch match (game 1 3), Seville 1 987.

ll .• .'il'xa2 12 0-0 (D)

B

Most of the games in the system with .l:tb1 pass through this position with its typical opening debate about which takes precedence: the superiority in the centre or the extra pawn. The main line is . . .

12 ... .i.g4 Before examining this move further, we shall

take a look at Black's other options in increasing order of popularity:

a) 12 . . . .i.d7 is a somewhat unexpected move, voluntarily returning the pawn at once, in order to vacate the square for the rook and quickly generate counterplay along the c-file. However, White's presence on the seventh rank has created the preconditions for several impressive routs; here is one: 1 3 .l:txb7 .l:tc8 (attacking the e4-pawn by 1 3 . . . .i.c6 leads to complications that favour White: 14 .l:txe7 �a3 15 .i.g5) 14 .i.g5 h6 1 5 .i.h4 (but not 15 .i.xe7? �e6) 15 . . . .i.c6 16 .l:txe7 .i.f8 17 d5 .i.a4 1 8 �d3 g5 1 9 .i.xg5 hxg5 20 .l:txf7 �xf7 2 1 lt:Jxg5+ �e7 22 e5 .i.c2 23 �f3 1 -0 Shipov-V.Ivanov, Petrosian Memorial, Moscow 1995.

b) The active-looking 12 ... lt:'lc6, inviting the d-pawn forward, is rarely seen, as the queen's way back home can become quite long, despite the fact that the g7-bishop is now included in her defence.

c) Out of several possible methods of developing the knight, the most common is the modest 1 2 . . . lt:'ld7. The knight is then transferred to b6, where it blocks the pressure along the b-file and controls important squares; on the other hand, it is removed from the attack on the white centre. After 1 3 .i.b4 lt:'lb6 14 .l:ta1 �e6 1 5 �b1 .i.d7 16 .l:ta5 Black must defend carefully.

d) Black can pin his hopes on his passed pawn and immediately set it in motion by 12 . . . a5 (D). Meanwhile, the queen remains in the danger zone.

w

d 1 ) The attempt to take advantage of that by 13 �c 1 is met with a counterattack in the centre: 13 ... .i.g4 (the evacuation by 1 3 ... �e6 looks riskier: 14 .i.c4 �xe4 and after 1 5 .i.xf7+ or 1 5 .l:te1 White wields a serious initiative) 14 .i.c4 �a4 1 5 .l:txb7 .i.xf3 1 6 gxf3 lt:'lc6 1 7 .i.b5 �xb5 1 8 .l:.xb5 lt:'lxd4. Now White didn't risk entering the line 1 9 �c5 lt:'lxf3+ 20 �g2 lt:'lxd2 2 1 .l:te1 a4, where he has extra material but Black has a strong passed pawn, and instead returned the queen by 1 9 .l:txa5 leading shortly to a draw in Kramnik-Leko, Dortmund 1996.

d2) 13 d5 is the other attempt to exploit the queen's position, by cutting off its line of retreat. Sharp play commences after 1 3 . . . a4 14 �c 1 b5 1 5 .l:txb5 .i.a6 16 .i.c4 .i.b2, trying to hold on with the help of tactics. The win of two pieces for a rook leads to an unclear game: 17 .i.xa2 .i.xc 1 1 8 .l:txb8 .l:tfxb8 1 9 .l:txc l .l:tb2 20 .l:ta1 ; in such a configuration the passed a-pawn is usually worth a considerable amount.


d3) 1 3 i.g5 (D) is the most unpleasant continuation for Black.

B

He has two ways to swap the e7 -pawn for the one on e4, but neither yields equality:

d3 1 ) 1 3 . . . a4 14 i.xe 7 l:te8 1 5 i.d6 l:.xe4 1 6 l:tc 1 lbc6 ( I 6 ... i.e6 17 i.d3) 1 7 i.c4 'ii'b2 I 8 l:tb1 'ii'c3 I 9 i.xf7+.

d32) I3 ... h6 places g5 under control at once, but then the queen is in danger, and in most lines the exchange has to be given up without sufficient compensation as well: 14 i.xe7 l:te8 1 5 i.d6 l:txe4 I 6 l:.c1 and now 1 6 . . . lbc6 1 7 i.c4 'ii'b2 1 8 l:tb I 'ii'c3 I 9 i.d5 l:.xd4 20 lDxd4 lbxd4 21 'ii'e 1 , 16 . . . 'ii'e6 I 7 i.d3 or 16 . . . i.g4 I 7 i.c4 'ii'b2 1 8 h3 ( 1 8 i.c5 ! ?) 18 . . . l:xd4 I 9 'ii'e 1 .

e) Immediately bringing the queen back by I2 ... 'ii'e6 leads to a less stormy course of events, but the queen is placed poorly on this square, gets in the way of everything else, and White systematically builds up the pressure. We shall give the game Lautier-Seger, Bundesliga 2002/3 as an example: 1 3 'ii'c2 'ii'c6 14 'ii'd3 ( I4 i.c4 is also possible) 14 . . . 'ii'd6 15 d5 lbd7 16 i.b4 lbc5 1 7 'ii'e3 b6 1 8 l:tfd1 i.g4 1 9 e5 'ii'c7 20 l:te1 l:tad8 21 l:tbd1 i.c8 22 l:td4 i.b7 23 i.c4 'ii'b8 24 h4. Having gained unequivocal superiority in the centre and tied Black's forces down, White prepares to open a second front on the kingside.

f) I2 . . . b6 is a solid continuation, preparing the natural development of the bishop and blocking the b-file. The queen will remain behind the enemy's lines for as long as possible to reduce White's freedom of action. This is an important idea relevant to many other such opening lines where Black spends time with his queen devouring one of White's queenside pawns: the queen can also have considerable

nuisance value, a factor which is squandered if the queen retreats unprovoked. White usually replies with 1 3 'ii'c I (D), not letting the bishop out on a6 and creating real threats to the black queen.

B

How Black countenances these threats will determine his reply:

fl ) He can retreat immediately by 13 . . . 'ii'e6. Then:

fl l ) For a long time the main continuation was 14 i.c4 with a subsequent sacrifice of exchange and a strong attack: 14 . . . 'ii'xe4 15 l:te1 'ii'b7 16 i.b4 i.e6 17 l:txe6 fxe6 1 8 lDg5. However, in the line 18 . . . lbc6 1 9 lDxe6 �h8 Black started to defend successfully.

fl2) Therefore the move I4 l:e1 came under scrutiny. However, here too, Black can hold on: I4 . . . i.a6 15 i.d I i.d3 I 6 l:tb3 i.c4 17 l:tc3 b5, Gladyszev-Nepomniashchy, Moscow 2006. We should also mention the game MalakhatkoWang Yue, Balagne 2007, featuring I4 . . . 'ii'xe4, a pawn-grab rejected by the theory. White reacted in a modest manner with I 5 lDe5 and got nothing after 15 . . . i.f5 I6 'ii'b2 lbd7. It would be interesting to know what Black had in mind in case of the main I 5 lbg5; after I 5 . . . 'ii'xd4 I6 i.f3 i.f5 I 7 l:tb4 'ii'd6 I 8 i.xa8 lba6 he recoups some of the losses, but is it enough?

f2) With 1 3 . . . i.b7, Black gets on with development. No mortal danger threatening the queen is apparent, so she stays in place to make White uncomfortable. This position serves as yet another platform for registering a draw, viz. I4 i.c4 'ii'a4 I5 i.b5 'ii'a2 with a repetition. If White has different requirements, he chooses I6 l:tei l:tc8 17 'ii'd I . Here 17 . . . e6 is sometimes played, when I 8 h4, suggested by Ivanchuk, is not bad. If the queen starts to get ready for a


return home with 1 7 ... 'ifc2, then 1 8 'ife2 'ifc7 ( 1 8 . . . lbc6 has been tested, when 1 9 i.d3 'ifa2 20 l:.a1 'ifh2 2 1 l:ta4 'ifb3 22 .:teal lbd8 23 l:t 1 a3 'ife6 24 l:txa7 allows White to win back the pawn while keeping a significant part of the accumulated positional pluses) 1 9 l:tbc1 'ifd8 20 l:txc8 'ifxc8 2 1 l:tc 1 'ii'd8 22 i.f4. On the whole, having avoided tactical skirmishes, White retains the strategic initiative.

We now return to 12 . . . i.g4 (D):

w

This bishop move is the central avenue of the system.

13 i.e3 When it became clear that Black is OK with

1 3 l:txb7 i.xf3 14 i.xf3 i.xd4 15 e5 (lately 1 5 i.b4 has been played, when some accuracy is required of Black in order to equalize) 1 5 . . . lba6, White turned his gaze to the d2-bishop.

However, instead of the text-move, 1 3 i.g5 looks the most active. However, the assault on e7 is not dangerous yet, and so Black can reply 13 . . . h6, without delay inviting the bishop to state on which diagonal he sees his future. Then:

a) Capturing the pawn by 14 i.xe7 did not remain in practice for long: 14 . . . l:te8 15 l:txb7 lbc6 16 i.c5 l:txe4 17 i.d3 l:txd4 1 8 i.xd4 lbxd4 1 9 i.e4 l:td8 20 'ii'b 1 lDe2+ 21 'ifi>h 1 'ifc4 with sufficient compensation, Kramnik-Kasparov, Novgorod 1 994.

b) 14 i.h4 poses more complex problems. Black has no time to defend b7, while . . . b6 dangerously weakens the long diagonal, so the correct course is not to cling to the extra pawn but to strive for active counterplay. The attack on the pawn-centre by 14 . . . g5 15 i.g3 lbc6 1 6 d5 l:tad8 17 l:txb7 f5 18 l:tc7 fxe4 1 9 l:txc6 exf3 20 i.c4 leaves White with the better chances, but 14 . . . a5, immediately setting this valuable pawn

in motion, is more interesting. After 1 5 l:txb7 g5 1 6 i.g3 a4 the strong passed pawn promises Black good prospects.

c) More often White prefers not to leave the c l -h6 diagonal and retreats with 14 i.e3, when there generally follows 14 . . . lbc6 15 d5 (D).

B

The difference between this position and the one in the main game is, obviously, the black h-pawn 's advance. The evaluation of such a difference is not straightforward, and in some lines it favours White, while in others his opponent. If Black carries out a plan similar to the one in this game, then after 1 5 . . . lba5 the bishop is deprived of the g5-square, which can be assessed as a feather in Black's cap. Thus White plays 1 6 i.c5, when in the variation 16 . . . i.f6, which for a time used to be the main line, 1 7 e5 i.xe5 1 8 l:tb4 i.xf3 19 i.xf3 l:tae8 20 i.e3 lbc4 21 i.xh6 allows White to take advantage of the h6-pawn's position. However, these days 16 . . . b6 is played, with the same ideas as in the game; then the placement of the h-pawn has no substantial significance. 15 ... lba5 is not the only move, but the game Bacrot-McShane, Biel 2004 cast doubt on the attempt to use the pin on the d-pawn: 15 . . . l:tfd8 16 lhb7 e6 1 7 l:tc7 i.xf3 1 8 i.xf3 lbe5 19 d6 i.f8 20 i.c5; on the other hand, 1 5 . . . i.xf3 16 gxf3 lDd4 17 i.d3 a5 1 8 f4 b5 leads to uncompromising play, with winning chances for both sides.

13 ... lbc6 14 d5 (D) 14 ... lba5 The availability of the move .tg5 works out

to White's advantage after 14 . . . l:tfd8, as well as in the line 14 . . . i.xf3 1 5 .txf3 lDe5 16 .te2 lt:Jc4 17 i.g5.

15 i.gS

MODERN EXCHANGE VARIATION WITH '!:.bl 81

B

The attack on e7 from this side looks better; at the very least 1 5 . . . .i.f6 is ruled out due to the loss of control over the at-square.

15 ••. b6 (D) Black can also defend the pawn, by 1 5 . . . 'ii'a3.

Thanks to his superiority in the centre, White retains compensation for it after 1 6 .i.d2 .i.c3 17 .i.e I 'ii'd6 1 8 eS or 1 6 .l:tel .i.xf3 1 7 .i.xf3

'i!kd6 1 8 'ii'e2 .i.eS 1 9 g3, although Black has his fair share of the play too.

w

16 .i.xe7 It is not apparent what else can be added, and

White embarks on the familiar operation involving the win of the exchange.

16 ••• .l:tfe8 17 d6 lLic6 18 .i.b5 l2Jxe7 19 h3 (D)

An attempt to improve on the game Gelfand-Van Wely, Wijk aan Zee 2006, where 1 9 .i.xe8 l:txe8 20 dxe7 'i!ka3 2 1 h3 .i.xf3 22 'ii'xf3 'ii'xf3 occurred, with a quick draw.

To untangle his pieces, Black must make a preparatory move; that's why White can keep the queens on the board, and he hopes that this will give him a chance to fight for an advantage.

19 ••• .i.xf3 20 'ilkxf3 'i!ke6 21 .i.xe8 .l:txe8 22 dxe7 .l:txe7

B

Thus White has won the exchange for a pawn, but on the other hand Black has a strong bishop and two passed pawns. They are not difficult to blockade, but White's advantage can be proven only by advancing his pawn-majority in the centre, and this is not easy to carry out and it involves risk for his own king.

23 :re1 .td4 The bishop is transferred to the ideal post on

cS, where it is protected, puts pressure on f2, keeps three files closed to the enemy rooks and supports the a-pawn in its desire to advance as far as possible. Black plans an exchange of rooks, as his own has no great role to play while White's attacking chances will be decreased.

24 .l:tbd1 'ii'e5 25 .l:td3 a5 26 'ii'd1 .i.c5 (D)

w

27 .l:te2?! 27 .l:tdS deserved consideration, taking ad

vantage of the fact that 27 . . . 'ilkf4 ?? is impossible due to 28 .l:txcS bxcS 29 'i!kd8+. The upshot is that White could have advanceo in the centre.

27 ••• .l:te6 28 g3 :d6 29 'iitg2 .l:txd3 30 'ii'xd3 a4 31 .l:td2 a3 32 'ii'c4 'iitg7 33 .l:td7 'ilkf6 34 f4

White has to make up his mind whether to advance or retreat. The king's position on h4 seemed to him secure enough.


34 ••• 'ii'h2+ 35 �f3 'ii'f2+ 36 �g4 h5+ 37 �h4 g5+ 38 fxg5 �g6 (D)

w

Oops ! Suddenly a terrible check on f4 is threatened, whereas the totally exposed black king is for the moment out of White's reach.

39 'ii'c3? 39 l:td3 ! holds: 39 . . . a2 40 'ii'd5 al'ii' 41 'ii'c6+

i.d6! 42 l:txd6+ �h7 ! 43 l:th6+ �g7 44 'ii'e8 'ii'xg3+ and Black gives perpetual check. But only a computer could be expected to find such a resource over the board, especially in a rapid game.

39 .•• f6?! (D) It appears that 39 . . . a2 ! is more precise: 40

l:td2 f6 is no good, while in the line 40 l:ta7 a l'ii' 4 I l:txal f6 42 gxf6 i.d4 43 l:tfl 'ii'xfl 44 'ii'xd4 b5 White, even though it is his tum to move, has to pick one of several lost pawn endings.

w

40 l:td5 The endgame after 40 gxf6 'ii'xf6+ 41 'ii'xf6+

�xf6 42 �xh5 b5 43 l:tc7 i.f8 ! (Nunn) should

be winning for Black, because his pawns are too fast.

40 •.• a2 41 l:tf5?! White misses his last chance: 4 1 l:txc5 bxc5

42 'ii'e5 - an attractive stalemate idea, isn't it? A study of the queen ending after 42 . . . fxg5+ 43 'ii'xg5+ would take too much space and not be relevant to the purpose of this book, so we shall stop at this point. At first sight, the black king escapes the checks, and the a-pawn must decide the game.

41 ••• 'ii'f4+ (D) 4 l ...a1 'ii'! ? is an even more effective version

of the same idea.

w

Both pretty and effective. The rest requires no commentary; the b6-pawn is protected indirectly, so White doesn't get to even dream about building a fortress.

42 gxf4 i.f2+ 43 'ii'g3 i.xg3+ 44 �xg3 al 'ii' 45 lbf6+ �g7 46 e5 b5 47 �h4 b4 48 �xh5 'ii'dl+ 49 �h4 b3 50 e6 b2 51 l:tf7+ �g8 52 l:tb7 bl'ii' 53 l:txbl 'ii'xbl 54 �g4 'ii'e4 0-1

Let's sum up the conclusions in the system with :bl . Black feels secure enough both in the 9 . . . ltJc6 variation with the subsequent retreat of the bishop to c7 (Game 1 5) and when he takes the a2-pawn (Game 16).

The general impression from the practice of the last few years is that White is getting fed up with such play, and today's fashion follows a different direction. However, there are still plenty of interesting possibilities for both sides, and the last chapter of the l:tb I story has definitely yet to be written.

6 Russian System : White Plays iVb3

Besides the most natural method of seizing the centre examined in the previous chapters, there is one that is more refined - to make the d5-pawn an offer to take on c4 that it cannot refuse, thereby removing the obstacles to the e4 advance. The first idea to present itself is increasing the pressure with the queen's leap to b3 - at once or after a preliminary check on a4. Both of these continuations are available after the development of the king's knight as well, so there are fewer moveorder issues if White wishes to make this his main weapon against the Griinfeld. Among these possibilities, centre stage is occupied by the variation 1 d4 lLlf6 2 c4 g6 3 lLlc3 d5 4 lLlf3 i..g7 5 'ii'b3 (D).

B

The defence by means of 5 .. . e6 involves significant positional concessions (a passive c8-bishop, the weakening of the f6-square) and is not seen in serious practice. 5 . . . c6 is more solid, even though it runs counter to the c-pawn's main functional task - to improve the g7-bishop's view out of the window by attacking the centre with . . . c5. Therefore Black usually doesn't get obstinate and gives up the centre by 5 . . . dxc4, hoping to gain compensation in better development by attacking the exposed white queen. After 6 'ii'xc4 0-0 7 e4 the main road parts into three branches:

7 . . . a6 prepares an extended fianchetto with the fianchetto of the queen's bishop as well. Game 17 (Bellaiche-Sutovsky) is devoted to this variation. In the notes we also cover the earlier options (including 'ii'a4+ ideas) for both sides up to move 7.

7 . . . i..g4, with the traditional Griinfeld counterplay against the d4-pawn, will be examined in Game 18 (Babula-Navara). This is called the Smyslov System and, as the name suggests, it has a rich heritage and was considered the main line in the 1 950s and 1 960s. Kasparov also used it in one of his matches with Karpov, but abandoned it following a heavy defeat. Black's ideas include . . . liJfd7 and . . . lLlc6, with piece-pressure on the centre coming before (or instead of) any pawnbreaks. The notes to this game also touch upon 7 . . . lLlc6, which can have ideas in common with the Smyslov System, but can also branch out in totally different directions.

Although strange at first sight, the flank development of the knight by 7 . . . lLla6 intends the rapid preparation of the freeing thrust . . . c5. While named after Prins, the variation was brought under the spotlight by Kasparov's adoption of it in the late 1 980s. This line is the subject of Game 19 (Nayer-Kumosov).


Game 1 7 [0971

Antony Bellaiche - Emil Sutovsky Catalan Bay 2007

1 d4 tLlf6 2 c4 g6 3 tLlc3 dS 4 ttJf3 As already mentioned, White has the option

of immediately increasing the pressure on d5 B by 4 'ifb3 (D).

B

If Black chooses to react in a standard fashion, then after 4 . . . dxc4 5 'ifxc4 ii.g7 6 e4 0-0, White, if he wishes to get something different rather than the main system, can choose to play 7 �e2 instead of 7 tLlf3. Now 7 . . . tLla6 pushes White back onto the beaten track (of Game 1 9) all the same; if that doesn't appeal to Black, then he can play in similar fashion to the Smyslov System (compare Game 1 8), except that, due to the knight's absence from f3, the undermining thrust . . . e5 is readily available: 7 . . . tLlfd7 8 i..e3 tLlb6 9 'ii'c5 e5. The attempt to put pressure on the white centre with the help of an immediate diversion on the queenside by 7 . . . c6 8 tLlf3 b5 9 ii'b3 'ifa5 is also interesting.

The preliminary check 4 1Wa4+ has the point of luring the bishop out to d7, where it blocks the d-file and leaves the b7-pawn without protection: thus 4 .. . i..d7 is met by 5 'ii'b3. Besides the attempt to reach familiar positions with 5 . . . dxc4, it is possible to take advantage of the d4-pawn's lack of protection and accept the pawn sacrifice with 5 . . . tLlc6 6 tLlf3 tLla5 7 1i'c2 dxc4, although White obtains good compensation.

4 ••• i..g7 5 'ifb3 With the knight developed to f3, defending

the d4-pawn, there are more reasons for 5 'ifa4+ (D), as this denies Black certain options.

In case of 5 . . . c6 White can reach a position akin to the Exchange Variation by 6 cxd5 tLlxd5 7 e4.

The principal line runs 5 . . . i..d7 6 'ii'b3 dxc4 (here there is no alternative) 7 'ii'xc4 0-0 8 e4. Now:

a) 8 . . . a6 is pointless here, as the bishop is far from the b7-square.

b) On the other hand 8 . . . b5 is possible now. Then 9 tLlxb5 tLlxe4 10 1i'xc7 (after 10 tLlxc7 tLlc6 1 1 tLlxa8 'ii'a5+ Black's initiative is too strong) I O . . . tLlc6 1 1 �d3 tLlb4 1 2 i..xe4 i..xb5 1 3 'ifxd8 l:.axd8 provides sufficient compensation for the pawn; Black has rich counterplay after 9 'ifb3 c5 as well.

c) 8 ... i..g4 is the Smyslov System (see Game 1 8), with only the move-number changed, as both sides have lost a tempo.

d) The choice of 8 . . . tLlc6 presupposes that Black is ready to play the game with a nonstandard material balance after 9 e5 �e6 (the bishop occupies the knight's natural retreatsquare, so the following operation is forced) 10 exf6 ii.xc4 1 1 fxg7 �xg7 1 2 i..xc4 tLlxd4 13 ttJxd4 'ii'xd4. As a rule, three minor pieces are stronger than a queen, but here Black has two extra pawns as well.

s ... dxc4 After 5 . . . c6 White can count on a small but

durable advantage by playing 6 cxd5. Then the recapture 6 ... cxd5 transposes into the Exchange Slav, where fianchettoing the dark-squared bishop is not the best decision, while 6 . . . tLlxd5

RUSSIAN SYSTEM: WHITE PLAYS "flb3 85

grants White a choice of various methods of obtaining superiority in the centre.

6 'ii'xc4 0-0 7 e4 The preliminary assault on c7 by 7 �f4

leaves Black with a choice: transpose into a sideline of the 7 . ..l'i:Ja6 system (see note 'e' to White's 8th move in Game 19), or play 7 . . . c6 with queenside pawn activity to follow.

7 ••• a6 (D) Here too, 7 . . . c6 is often employed. Besides

supporting the possible raid by the b-pawn, this move takes control over the d5-square, which can be significant if Black switches to undermining the centre from the other side. Here is a fresh example: 8 'S'b3 e5 9 dxe5 ltJg4 10 �e2 ltJxe5 1 1 lL'lxe5 .txe5 1 2 0-0 Ve7 1 3 �e3 �e6, Radjabov-Mamedyarov, Wijk aan Zee 2008.

w

The text-move was first developed by Hungarian players about forty years ago. Black will follow up with ... b5 in almost all cases, with . . . �b7 and . . . c5 also very much on the agenda.

8 �e2 White develops, for the moment keeping the

pawn configuration in the centre intact. But before moving on to an examination of

this move, we should consider the fundamental basis for Black's little move with his a-pawn:

a) First of all, Black agrees to sacrifice a pawn, for which he will obtain sufficient compensation - 8 .tf4 b5 9 Vxc7 'ii'xc7 10 �xc7 �b7 1 1 e5 ltJd5 1 2 ltJxd5 �xd5 with the subsequent . . . ltJc6 and . . . l:r.fc8.

b) We should note that White cannot prevent . . . b5 by 8 a4? ! since then 8 . . . b5 follows anyway, because 9 axb5?? axb5 is a catastrophe for White because of the loose rook on a l .

c ) 8 'S'b3 i s an important possibility. After 8 .. . b5, 9 e5 transposes to line 'd' , while 9 .te2

leads to the position seen in the main game. One independent continuation is 8 . . . c5 9 dxc5. Black wins the pawn back after 9 ... Va5 10 'ii'b6 'S'xb6 I I cxb6 ltJbd7 or 9 . . . ltJbd7 10 c6, and in the subsequent manoeuvring struggle White seeks an advantage based on his slightly better pawn-structure.

d) The principal alternative to the approach seen in the main game is 8 e5 b5 9 'ii'b3 (D).

B

White grabs space, and will often put this to use to launch a direct attack, with the move e6 a factor in many lines.

d l ) 9 . . . ltJg4?! is not so common, since after 10 h3 ltJh6 it is hard for Black to destroy the white centre. White can consolidate by 1 1 �f4 with a subsequent l:r.dl or first exchange off Black's active light-squared bishop by 1 1 �d3 �b7 1 2 �e4.

d2) A modification of a previously-seen idea is clearly unfavourable here: 9 . . . �e6? 10 exf6 �xb3 I I fxg7 �xg7 1 2 axb3.

d3) 9 . . . ltJfd7 is the standard move, when White has three plans:

d3 1 ) 10 h4 immediately launches a standard attack, exploiting the temporary absence of the f6-knight. In full accordance with the principle that a blow in the centre is an effective counter to flank operations, Black replies IO . . . c5 ( 10 . . . ltJb6 I I h5 c5 1 2 hxg6 hxg6 1 3 dxc5 �e6 14 'ii'c2 �f5 looks reliable as well; in order to avoid a repetition, the queen has to head for a3 or e2). Now the e5-pawn is doomed and tries to sell its life for maximum profit by 1 1 e6 (D).

Now Black must make a fundamental decision. White's centre can be reduced to rubble by l l . . .fxe6 12 h5 cxd4, but 1 3 hxg6 ltJc5 14 gxh7+ �h8 1 5 ltJh4 gives White a potent attack. It is no less dangerous if Black begins with


B

l l . . . cxd4 1 2 exf7+ (the win of a piece for two pawns by 1 2 exd7 hands Black the initiative: 1 2 . . . lt:Jxd7 13 lt:Je4 i.b7 14 i.d3 i.xe4 1 5 ..txe4 lt:Jc5 1 6 'ili'c2?! l:!.c8) 1 2 . . . �h8 1 3 h5. The other option is l l . . .c4, to neutralize White's activity with rapid development and piece control of the centre. Black not only gains a tempo but the d3-square as well, and the retreat to c2 will present in the future another tempo for the bishop on f5. After 1 2 'ili'dl lt:Jb6 ( 1 2 . . . lt:Jf6 blocks both the f-file and the bishop) 1 3 exn+ l:!.xf7 14 h5 lt:Jc6 1 5 hxg6 hxg6 16 i.e3 i.f5 17 lt:Jg5, the white king is also not safe, and there is counterplay after both the exchange sacrifice 17 . . . tt:Jxd4 1 8 tt:Jxn �xf7 1 9 l:!.c 1 e5 and 17 . . . l:!.f6 1 8 g4 i.e6 1 9 lt:Jce4 i.d5 20 f3 e5, Kasimdzhanov-Khalifman, Wijk aan Zee 2002.

d32) White can begin with 10 e6, while there is no threat driving the queen off b3. It is undesirable to allow the pawn to f7, so Black replies 10 . . . fxe6. Taking on e6, even with check, will eventually prove a loss of time, so White continues his development, simultaneously hindering . . . c5 : 1 1 i.e3 (D).

B

If undermining the centre doesn't work, it is time to relocate the knight from d7. It can

choose l l . ..lt:Jb6, when White returns to the plan of the kingside attack by 1 2 h4 lt:Jc6 1 3 h5. Practice has shown that the exchange sacrifice 13 . . . l:!.xf3 14 gxf3 lt:Jxd4 1 5 l:!.d1 is insufficient for Black, so the knight started to prefer the other direction: l l . . .lt:Jf6. In reply to this it is logical to create weaknesses on the opposite flank by 1 2 a4, but Black returns the pawn with 1 2 . . . b4 1 3 'ili'xb4 lt:Jc6 and obtains counterplay along the b-file.

d33) 10 i.e3, hindering . . . c5 and leaving the e6 advance in reserve, is the most natural continuation. During the initial stage of the variation's development Black used to choose 1 0 . . . lt:Jb6, not allowing the e6 advance, yet also giving up . . . c5. In that case mutual weaknesses appear, first of all Black's c5-square and the cpawn, and the d5-square in the opposite camp; the game takes on a manoeuvring aspect, although White tends to have the lead. Then the thematic 10 . . . c5 became the main reply, inviting the 'kamikaze' forward: 1 1 e6. After l l . . . c4 there arises a position similar to line 'd3 1 ' , with the difference that the bishop is on e3 instead of White having played h4. This means that the white king's residence on the kingside is still safe, and White can initiate play on the opposite side by means of a4. Later Timman introduced 1 1 . . .cxd4 into practice. Black can react to the win of a piece by 1 2 exd7 in two ways: 12 ... 'ili'xd7 1 3 l:!.d1 e5 14 .te l i.b7 or 12 ... dxe3 1 3 dxc8'ili' exf2+ 14 �xf2 'ili'b6+, in both cases obtaining two pawns and a strong initiative. If 1 2 exf1+ l:!.xf7 13 i.xd4 ( 1 3 lt:Jg5 is well met with l � . . . lt:Je5), then after 13 . . . i.xd4 14 lt:Jxd4 lt:Jc5 1 5 'ili'd5 'ili'xd5 16 lt:Jxd5 lt:Jbd7 the active pieces compensate for the extra pawn-island.

8 ••• b5 9 'ili'b3 c5 (D) Also possible is 9 ... i.b7 10 e5 lt:Jd5 1 1 0-0

c5 12 dxc5 lt:Jd7 13 lt:Jxd5 lt:Jxc5 14 lt:Jxe7+ 'ili'xe7 15 'ili'a3. Black will regain the pawn by 1 5 . . . l:!.ac8 16 i.f4 I:tfe8 17 l:tac 1 i.xf3 1 8 i.xf3 i.xe5, but the bishop-pair promises White a slight initiative.

After the text-move, White has to bid farewell to the centre and prove that the passed c5-pawn is an advantage in a different form.

10 dxc5 ..tb7 If the bishop picks a different workplace by

10 . . . i.e6, then Black wins back the c-pawn and with accurate play can hope for equality: 1 1

RUSSIAN SYSTEM: WHITE PLAYS 'iflb3 87

w

�c2 tDc6 1 2 0-0 �c7 1 3 h3 tiJb4 14 'iWb 1 'ii'xcS 1S .te3 �c7.

11 0-0 ltJxe4 12 ltJxe4 .txe4 13 l:ld1 This move has been played a few times but is

not considered to be dangerous. In order for the passed cS-pawn to become a

real factor, White needs to complete his development and rearrange his forces without dropping the b2-pawn, unless he obtains strong compensation, as the black bishops keep the queenside under dangerous fire. The development of the dark-squared bishop suggests itself, as then the queen can take up a position on e3 with gain of tempo, and the exchange on h6 can be offered. Thus:

a) After 13 .tgS tDc6 14 �e3, one possibility for Black is 14 . . . .tfS l S .ih6, when the unclear exchange sacrifice 1S . . . .txb2 16 l:lad1 �aS 1 7 .txf8 l:lxf8 has been tested. Later another continuation was recognized as more reliable: 14 . . . 'ii'dS I S l:ladl 'ii'e6 1 6 .i.h6 .tc2 with a forced queen exchange.

b) Today the search for an advantage is focused on another bishop move, 13 .tf4 (D).

B

Here e7 is not under attack, and Black can inquire at once into White's plans as regards the

b2-pawn by 1 3 . . . .tdS. A critical line seems to be 14 'fl'e3 .txb2 1 S l:tad1 e6 1 6 ltJeS. In Onishchuk-Svidler, Foros 2007, Black immediately parted with the bishop by 1 6 . . . .txeS 17 .txeS tDc6 and held out, but further developments can be expected here.

13 ... 'f:i'c7 14 .ie3 ttJc6 15 l:.d2 Usually I S a4 has been played here. Instead

White defends b2 and wants to prepare the exchange of the dark-squared bishops but loosens the grip on the d4-square.

15 ..• e6 16 .tgS The attempt to exchange off the other bishop

brought White nothing good in Tisdaii-Korchnoi, San Francisco 1 99S: 1 6 ltJgS .idS 1 7 �c2 l:ad8 1 8 .tf3 tiJb4 1 9 �d 1 .txf3 20 'fl'xf3 tiJdS 2 1 l:ad I tDxe3 22 'f:i'xe3 .txb2.

16 .•• h6 17 .th4 .txf3 18 .txf3 liJd4 19 �e3 (D)

B

19 ... ttJr5 Black had to choose between this move and

19 . . . tDxf3+ 20 gxf3. 20 .tg3 eS In spite of the shattered pawns, White is OK

in the endgame arising after 20 . . . 'f:i'xg3 2 1 hxg3 tDxe3 22 fxe3, since the strong passed pawn provides sufficient counter-chances. However, the text-move doesn't promise Black much either.

21 1i'a3 l:tad8 22 l:lad1 l:lxd2 23 l:lxd2 l:ld8 24 l:d5?!

Unnecessary complications. After 24 l:lxd8+ 'f:i'xd8 2S c6 lDxg3 26 hxg3 e4 27 .txe4 'f:i'd1 + Black has a draw for the asking, whereas attempts to play for a win are already dangerous. It seems that White, too, dreams of more.

24 ... ttJd4 25 .l:.xd8+ 'i:i'xd8 26 'fl'xa6 e4 27 .td1 tiJfS?!


The natural 27 . . . 1!i'g5 gives Black a strong initiative. The decentralization of the knight returns the game to the drawing zone.

28 ..td6?! Simpler is 28 ..tb3 and the queen can return

to the defence: 28 . . . 1!i'd3 29 11i'a8+ ..tf8 30 1!i'd5. 28 ••• e3 29 fxe3?! 29 11i'xb5 exf2+ 30 'iitxf2 looks more rational,

and the king escapes the house arrest, when the fact that White has an entire extra flank obliges Black to be careful not to push too hard.

29 ... tt'lxe3 (D) 30 ..tf3? This loses, since the bishop occupies the

square to which the queen had to return. However, it is quite likely that even after 30 ..te2 1!i'h4 3 1 1!i'a8+ Wh7 32 g3 1!i'd4 the search for salvation may tum out fruitless.

w

30 ... 1!i'h4 31 i.g3 Black also has a decisive attack in case of 3 1

g3 1!i'b4 32 1!i'a8+ 'iith7 3 3 h4 1!i'e1 + 34 'iith2 11i'f2+ 35 'iith3 g5.

31...1!i'c4 32 h4 1!i'cl + 33 'iith2 tt'lfi + 34 'iith3 tt'lxg3 35 'iitxg3 1!i'e1 + 0-1

Game 1 8 [099]

Vlastimil Babula - David Navara Czech Ch, Luhacovice 2003

1 d4 tt'lf6 2 c4 g6 3 tt'lc3 d5 4 tt'lf3 ..tg7 5 1!i'b3 dxc4 6 1!Vxc4 0-0 7 e4 i.g4

This system bears the name of Smyslov, who laid the first bricks of its foundation. Let's examine another possibility - the natural developing move 7 . . . tt'lc6 (D), also frequently seen in contemporary practice.

w

Black generally intends to reach the 'Smyslov' set-up via a different move-order, and in the meantime the g4-square remains accessible to the knight too. All reasonable replies have been tried. The knights' somewhat defiant position provokes the centre into advancing. The

d4-pawn is best advised to stay put since after 8 d5 tt'la5 9 1!i'd3 c6, White would be prudent to acquiesce to equality by 10 dxc6 tt'lxc6 1 1 1!i'xd8 while the option is there. Its colleague on e4 has better grounds for an advance: 8 e5. In the previous game we examined a similar position with the bishop on d7, where it was lured by the check on a4. That meant that the knight was deprived of this retreat-square, and Black was left with a choice between moving to an inferior square and the bishop move to e6 with the exchange of three minor pieces for the queen and two pawns, which is what Black usually did. 8 . . . i.e6 is possible here as well, but 8 . . . tt'ld7 is now available as an alternative, and is considered preferable. A possible course of events is 9 i.e3 tt'lb6 10 1!i'c5 a5 1 1 i.e2 tt'lb4 1 2 0-0 c6, with strong and weak squares on both sides and White enjoying a bit more freedom of action.

White can rule out . . . ..tg4 with the prophylactic 8 h3, but Black is OK then. Black has wasted no time and played actively, which cannot be said about the expedition by his opponent's queen; the standard transfer of the knight to b6 with a subsequent . . . f5 gives Black an excellent game.

RUSSIAN SYSTEM: WHITE PLAYS 'fib3 89

The development of the dark-squared bishop by 8 i.e3 elicits the natural reply 8 . . . l£lg4 with a further . . . e5, so the preliminary 8 i.f4 has been w tried, threatening the c7-pawn and intending to drop back to e3 only after 8 .. . l£lh5, but follow-ing 9 . . . i.g4 the knight turns out to placed rather well, eyeing the f4-square.

8 i.e2 (D) is considered to be the most solid continuation.

B

The pawn sacrifice 8 . . . e5 9 d5 l£Jd4 is inadequate: 10 l£Jxd4 exd4 1 1 'i'xd4 c6 1 2 'i'c4. Black should content himself with the fact that White has committed to developing this bishop to e2, and return into the main system of development with 8 . . . i.g4. He is agreeable to 9 ..ie3 i.xf3 and now 10 gxf3 e5 1 1 d5 l£Jd4 or 10 i.xf3 e5 1 1 d5 l£Jd4 1 2 i.d 1 a6 (or 1 2 . . . c6). For White, the search for an advantage has focused on the line 9 d5 l£la5 10 'i'b4 ..i.xf3 1 1 i.xf3 c6 1 2 0-0. Besides the traditional set-up 12 . . . cxd5 1 3 exd5 l:tc8 with the transfer of the blockading knight to d6, 1 2 . . . 'i'b6 is under development at the time of writing; e.g., 1 3 'i'a4 (the win of a pawn by 1 3 'iixe7 grants Black active pieces and good prospects of equality after 1 3 . . . l£lc4 14 dxc6 bxc6 1 5 i.f4 l:tfe8 16 'iic7 'iixc7 17 i.xc7 l:tac8 1 8 ..ig3 l£lh5) 13 . . . l£Jd7 14 l:.d 1 l:tfd8 1 5 'iic2 l:tac8 1 6 ..ie3 'i'b4 1 7 dxc6 bxc6 1 8 ..ie2 l£lc4 with mutual chances, OnishchukAvrukh, Biel 2007.

We now return to 7 . . . ..tg4 (D): 8 i.e3 This is considered to be the most unpleasant

continuation - White strengthens the centre and clears the way for the rook. Other moves:

a) 8 l£le5 suggests itself but has its drawbacks; after 8 . . . ..ie6 9 d5 ..ic8 the changes in the position consist of the advance of the pawn

and the knight. However, the pawn is easier to undermine on d5, and the unprotected steed requires a close watch; after I 0 ..ie2 e6 I I 0-0 exd5 1 2 exd5 l£le8 the transfer of the knight to d6 gives Black a pleasant game.

b) Less tense for Black is 8 i.e2 l£lfd7 9 i.e3 l£lb6, when the d4-pawn's lack of protection deprives the white queen of the retreat to b3. After 10 'iid3 l£lc6 1 1 l:td 1 Black has a choice: 1 l . ..i.xf3 1 2 i.xf3 e5, 1 1 . . .'iid7 or 1 1 . . .'iic8; for example, 1 1 .. .'iic8 1 2 'iic2 l:td8 1 3 d.5 i.xf3 14 i.xf3 l£Je5 15 i.e2 c6 16 'iib3 l£Jg4, Ki.Georgiev-Kamsky, World Cup, KhantyMansiisk 2007. 10 'iic5 also presents Black with a tempo, and IO . . . l£18d7 gives him good chances.

8 •.• l£lfd7 This move looks strange, as Black spends

time decentralizing an already-developed piece, but precisely this move is the main idea that the system is based upon. The functionality of the pieces is more important than good looks, and the b6-square is preferable to f6 due to the concrete specifics of the position. The idea is to cause White problems by attacking the d4-pawn (this theme started with the move . . . ..ig4), and the knight's manoeuvre unmasks the g7-bishop and prepares to attack the white queen, which is one of the d4-pawn's defenders. At an appropriate moment in the near future, Black will play . . . l£lc6, whereupon White will need to make some sort of concession to hold his centre.

For a long time the move-order 8 . . . l£lc6?! has been regarded as dubious: 9 d5 ..ixf3 10 gxf3 ( 1 0 dxc6 b5 !) I O . . . l£le5 1 1 'iie2 with f4, ..ig2 and 0-0 to follow, Botvinnik-Smyslov, Groningen 1 946.

9 'iib3

90 CHESS EXPLAINED: THE GRUNFEW

By retreating to b3 at once, White encourages the enemy knight not to get distracted from the b6-square. Naturally, there are several B other possibilities.

Considering that in the coming struggle the white king often remains in the centre, 9 0-0-0 (D) suggests itself, sending it to the queenside with no loss of time compared to playing .l:.d I .

B

But here too, it is insufficiently secure, as the g7-bishop exerts a strong influence on this side of the board, and the rooks have better chances of opening lines. Black has several promising plans: the pawn advance . . . c6 (or . . . a6) and . . . b5, or the attempt to clear the diagonal with . . . e5 or . . . c5; for example, 9 . . . c5 10 dxc5 'ifa5 1 1 'it>b I ltJc6 1 2 .i.e2 b5 13 cxb6 .i.e6 with a fun game, Piket-Shirov, Amber Rapid, Monaco 1 996. For his part, White, having a certain superiority in the centre, strives to simplify the position and obtain an advantage in the endgame.

The main alternative to the move in the game is 9 .l:.d I . Often this comes down to a mere transposition of moves, but independent lines appear after 9 . . . ltJc6, intending a different use for the d7-knight:

a) I 0 'ifb3 e5 I I dxe5 ltJcxe5 12 .i.e2 .i.e6. If White takes the b7-pawn, Black acquires an active rook on b2, while 1 3 'ifc2 ltJxf3+ and 14 . . . c6 is also OK for Black.

b) There are more branches in case of 10 .i.e2:

b l ) The line 10 . . . ltJb6 I I 'ii'c5 (D) has a high pedigree.

Let's recall a celebrated encounter: 1 1 . . . 'ii'd6 1 2 h3 .i.xf3 1 3 gxf3 .l:.fd8 14 d5 ltJe5 15 ltJb5 'iff6 16 f4 ltJed7 1 7 e5 'ifxf4! was played in Botvinnik-Fischer, Varna Olympiad 1962, where

the events in the opening turned out to be a mere hors d'oeuvre, and the subsequent endgame was analysed for decades. We should also mention the 1 986 match between Karpov and Kasparov, where the move 1 2 e5 was twice the subject of discussion. Today Black has no fear of this position.

b2) The voluntary capture 10 . . . .i.xf3 is occasionally seen. If White recaptures with 1 1 .i.xf3, then l l . ..e5 looks good, while in case of 1 1 gxf3 lDb6 1 2 'ifc5 the attempt to undermine the centre with the other pawn, i.e. 1 2 . . . f5, has been tested in recent practice. As an example, I would like to offer the reader the following miniature: 1 3 h4 f4 14 d5 ltJe5 1 5 .i.d4 'ii'd7? ! 1 6 ltJb5 .:.fc8 17 'ifc3 ltJa4? 1 8 'ii'b3 (it transpires that White has another threat as well: ltJxc7) 1 -0 Landa-Van der Weide, Hamburg 2007.

9 ..• ltJb6 White's last move loosened his control over

c5 and failed to bring a rook to the d-file. Black can try to take advantage of these facts with 9 . . . c5 (D).

w

White is offered two pawns to choose from, but he usually declines, as 10 dxc5 'ii'a5 or I 0

· RUSSIAN SYSTEM: WHITE PLAYS 'flb3 91

1i'xb7 .i.xf3 1 1 gxf3 cxd4 is quite fine for Black. There remains 10 d5, with 10 . . . ltJa6 as the reply. Initially, White sought to refute the variation by exploiting the exposed position of the g4-bishop after 1 1 tiJd2, but Black's counterplay after l l . ..e6 is considered adequate: 1 2 d6 (Black feels confident i n sharp continuations like 1 2 .i.c4 exd5 1 3 .i.xd5 ltJb4 or 1 2 h3 exd5 13 exd5 .i.f5 14 g4 c4) 12 . . . .i.d4 13 lDb5 e5 14 h3 .i.e6 15 .i.c4 .i.xc4 16 ltJxc4 1i'h4 with mutual chances, Lugovoi-Klimov, St Petersburg Ch 2004. Nowadays 1 1 .i.e2 is preferred, with positions close to those examined in the next game; the difference is that it is harder for Black to include the exchange of the e-pawns on d5, so he seeks counterplay after l l . . . l::tb8 or 1 1 . . .1i'b6.

10 l::td1 e6 It appears that I O . . . ltJc6?!, which for a long

time was the main move, is suffering from a serious illness. Events usually develop as follows: 1 1 d5 ltJe5 1 2 .i.e2 ltJxf3+ 1 3 gxf3 .i.h5 (D).

w

The post on h3 often turns out to be a deathtrap after f4-f5, but on h5 Black has to be alert as well. 14 f4 used to be popular, playing for an attack, but it leads to rather double-edged positions. Today White favours 14 l::tg l 1i'd7 (or 14 . . . 1!kc8) 1 5 l::tg3 c6 16 a4 1!kc7, combining play in the centre with threats on the kingside and against the h5-bishop.

In order to reduce the effect of the d5 advance, it is sensible to play .. . e6, although this presents White with a tempo for development. Black must decide whether to include the exchange on f3. If he is going to do so, now is the time, while White has no choice about how to recapture: 10 . . . .i.xf3 1 1 gxf3 e6. White can

play in the centre or against the king. In the first case he aims for an advantage in the endgame, not hesitating to sacrifice a pawn: 1 2 d5 exd5 1 3 ltJxd5 ltJxd5 14 lhd5 1i'f6 1 5 f4 1i'xb2 16 1i'xb2 .i.xb2 1 7 l::tb5 .i.c3+ 1 8 'it>e2 b6 1 9 e5 ltJa6 20 .i.g2 .:.ab8 2 1 l::tc 1 .i.b4 22 l::tc6, Beliavsky-Krasenkow, Barcelona 2007. The prospects of the attack are attractive as well: 1 2 h4 ltJc6 13 e5 h5 14 .i.g2 ltJe7 1 5 .i.g5 1!i'd7 16 f4 c6 17 ltJe4 1i'c7 1 8 .i.f3 ltJbd5 1 9 l::td2, freeing the square for the queen i n order to prepare the sacrifice on h5, M.GurevichKrasenkow, Jakarta 1 996.

11 .i.e2 ltJc6 (D) Here 1 1 . . ..i.xf3 1 2 .i.xf3 ltJc6 1 3 e5 looks

even more suspect, as White doesn't have to pay for the powerful light-squared bishop with the weakening of his pawn-structure.

w

12 ltJg1 This is playable thanks to the tempo spent on

. . . e6. Not being forced to advance, White keeps the mobility of the centre intact. A step forward by each pawn has been tested as well: 1 2 d5 simplifies the position and so reduces Black's problems, so 12 e5 has been played more frequently. Naturally, this weakens the d5-square, but we can also list the pros: the dark squares are fixed, and the g7-bishop becomes passive, while the knight acquires the e4-square and a path to c5. A manoeuvring struggle with chances for both sides lies ahead.

12 ... .i.xe2 13 ltJgxe2 1!ke7 The exchange that took place has a positive

aspect for Black as well - besides the fact that there are now fewer pieces on the board, White's control over the c4-square has been weakened, and Black can hope to sink a knight into it.

14 a3


The queen is contemplating a walk to b4, but at the moment White is not obliged to prevent it. After 14 0-0, 14 .. .'ii'b4?! is dubious in view of 1 5 'ii'xb4 tbxb4 1 6 d5 threatening .tc5, so the usual reply is 14 . . . l:tfd8, which we shall illustrate with the following game: 15 e5 h6 (without a knight on e4 it is hard for White to create threats, and with his reply Black forestalled a possible .tg5) 1 6 t"Lle4 'ii'b4 1 7 'ii'c2 'ii'c4 18 'ii'xc4 tbxc4 1 9 l:tc 1 t"Llxb2 20 l:tb1 tbc4 2 1 l:txb7 with the initiative in the endgame, Beliavsky-Ftacnik, Turin Olympiad 2006.

14 ... tba5 15 'ii'b5 tbac4 16 .tg5 'ii'e8 17 'ii'b3

After the exchange of the queens, no advantage is apparent either.

17 .•• e5 There have been other plans of counterplay

as well, such as 1 7 . . . a5, intending to fix the light squares. Black chooses not to include a preliminary 1 7 . . . h6, hoping instead to take advantage of the drawbacks of the bishop's position on g5.

18 d5 The complications after 18 t"Llb5 exd4 or 1 8

tbd5 lLla5 are O K for Black. 18 .•. f5 (D)

w

19 0-0 The exchange 1 9 exf5 gxf5 gives Black a

mobile pawn-pair and additional active possibilities.

19 .•• f4 20 f3 l:tf7 21 l:td3 .tf8 After a relatively quiet manoeuvring open

ing, the game enters a dynamic phase. The subsequent play resembles sharp lines of the King's Indian Defence. The difference is that the centre is not closed here, which makes it possible for Black to activate his 'bad' bishop,

while White obliges his opponent to take into consideration the already-exposed weakness on c7.

22 l:tcl .tc5+ 23 'iii>h1 .te3 24 .l:.c2 h6 White needs just one more move to encircle

Black's active pieces. A game of calculation begins, with the motto "whoever is first, wins!"

25 .th4 g5 26 .tel g4 27 tbd1 ?! 27 fxg4!? looks risky because 27 ... f3?! gives

Black an extremely strong attack in lines like 28 gxf3? l:txf3 29 'Ot>g2 'ii'f7 30 li:ld I l:tfl 3 1 .tg3 'ii'f3+ 3 2 'it>h3 l:txdl 3 3 l:txdl h 5 34 gxh5 'Ot>f7 35 tbg1 'ii'xh5+ 36 'it>g2 'ii'xd1 37 l:tf2+ .txf2 38 'ii'xdl tbe3+. However, 28 tbg3 ! is much better, because after 28 . . . f2 29 .txf2 l:txf2 30 l:txf2 .ixf2 3 1 tbf5 it is now White who has a powerful attack.

27 .•• gxf3 28 gxf3 l:tg7 (D)

w

29 tbxe3 There was a mating threat in the form of

... 'ii'h5, so the exchange sacrifice is forced. Nevertheless, White obtains compensation in the form of a pawn and Black's weaknesses on c7, e5 and h6, and this is sufficient to maintain the balance.

29 .•• tbxe3 30 l:txe3 fxe3 31 tbg3 'it>h7 32 tbf5 'ii'h5 33 'ii'xe3 tbxd5

Black decides to return the exchange, hoping to get an attack with material being even. The other method, 33 .. JH7 34 .td2 l:txf5 35 exf5 t"Llxd5 36 'ii'e4 c6 37 'ii'g4, doesn't promise much either.

34 'ii'd3? ---

A mistake, and one that is immediately fatal. 34 'ii'f2 l:tg5 35 t"Llg3 l:txg3 36 'ii'xg3 l:tg8 37 'ii'f2 tbf4 38 l:txc7+ 'it>h8 39 'ii'fl l:.g2 40 h4 holds the game.

34 ••. lLlf4 35 li:lxg7 'ii'h3 0-1

RUSSIAN SYSTEM: WHITE PLAYS "iVb3 93

Game 1 9 [097]

Evgeny Nayer - Igor Kurnosov Russian Ch, Krasnoiarsk 2007

1 d4 lLlf6 2 c4 g6 3 lLlf3 i..g7 4 lt::lc3 dS 5 'ii'b3 dxc4 6 'ii'xc4 0-0 7 e4 lLla6 (D)

w

Taking advantage of the fact that the queen obstructs the bishop's path to a6, Black brings out the knight to this square. Development to the edge of the board is condemned by most beginners' manuals, but there are no rules without exceptions, and as time passed, the value of a move began to measured as the sum of the accompanying virtues and drawbacks. In this case the knight supports the most effective attack upon White's centre, i.e . . . . c5, without getting in the way of the other pieces at the same time.

8 i..e2 The most natural and popular continuation,

though not the only one, of course. Let's examine the other possibilities:

a) 8 'ii'b3 clears the diagonal for the bishop, creating i..xa6 ideas. However, Black has time to strike in the centre by 8 . . . c5. Besides 9 d5 no good reply is apparent, and then Black needs to make a decision. He is not too worried about the spoiling of his pawn-structure, as White is not well enough developed to fix these weaknesses, while two good bishops and the b-file provide sufficient counterplay. To forestall a possible advance in the centre, he carries out the thematic 9 . . . e6, when a possible continuation is 1 0 i..xa6 bxa6 1 1 0-0 exd5 1 2 exd5 and now 1 2 . . . i..f5 or 1 2 . . .'i!i'b6.

b) 8 'ii'a4 puts the queen on a different square but has similar ideas. The assessment doesn't

change substantially, but there are certain differences after 8 . . . c5 9 d5. First, there is 9 . . . i..d7, but the exchange of the bishops with 1 0 i..b5 is more likely to favour White, as the undermining thrust . . . e6 is deprived of support. Second, after 9 . . . e6 10 i..xa6 bxa6 there are more reasons to choose the 1 1 dxe6 i..xe6 (not attacking the queen) set-up, agreeing to switch to defence for the time being.

White can begin with the development of the dark-squared bishop:

c) 8 i..e3 is rarely seen as it fails to hinder 8 . . . c5. After 9 dxc5 lLlg4 10 i..d2 i..e6 1 1 'il'b5 i..d7 1 2 'il'xb7 lt::lxc5 1 3 'il'b4 'il'b6 Black's activity compensates for the lost pawn.

d) By choosing 8 i..g5 c5 9 d5 White intends to make use of the pin after 9 . . . e6 10 0-0-0; for example, 10 . . . 'il'a5 1 1 'il'b5 'il'xb5 12 i..xb5 lt::lc7 1 3 .ie2 exd5 14 exd5 lld8 1 5 d6 lt::lce8 16 i..b5, Azmaiparashvili-Carlsen, World Cup, Khanty-Mansiisk 2005. However, Black can play better: 9 . . . h6 (the bishop must be removed from the diagonal) 10 i..h4 b5 1 l lLlxb5 'il'a5+ 1 2 lt::ld2 l:tb8 with a dangerous initiative for the pawn.

e) The main rival to the text-move is 8 i..f4 (this position is frequently reached via 7 i..f4 lt::la6 8 e4). As always, Black counters with 8 . . . c5 (D). Then:

w

e I ) The long search for an advantage after 9 dxc5 failed to bring White success. In reply, Black has two options:


e l l ) 9 . . .'�a5 is met by 10 e5, seeking to disturb the harmonious arrangement of Black's forces. 10 . . . ltJd7 1 1 a3 (forcing Black to capture the pawn with his queen, but also weakening his light squares) l l . . .'i!hc5 . Initially 1 2 ltJd5 was tested, until it transpired that the pawn sacrifice 1 2 . . . ltJb6 1 3 'ifxc5 tDxc5 14 ltJxe7+ �h8 gives Black good counterplay. White then switched to 1 2 'iVe4, but has yet to achieve consistent success. For example, Black can offer an exchange of queens by 1 2 . . . 'ifc6.

e l 2) The pawn sacrifice 9 . . . .i.e6 10 'iVb5 .i.d7 is no less tempting. After 1 1 'fixb7 ltJxc5 1 2 'ii'b4 ltJe6 1 3 i.e5 a5 14 'i!Va3 'ii'b6, 1 5 i.c4 can be met with 15 . . . l:tfc8, as played by Kasparov, while 1 5 . . . ltJg4, as advocated by Sutovsky, is also interesting.

e2) White can of course push forward: 9 d5 e6. Then I 0 d6 is of independent significance. We shall illustrate the possible course of events with an example from the work of highly esteemed players: 10 . . . e5 ( 1 O . . . ltJb4 and 1 O . . . ltJh5 have also occurred) 1 1 i.xe5 ltJb4 1 2 l:td l i.e6 1 3 1i'xc5 ltJc2+ 14 �d2 ltJd7 1 5 1i'c7 ltJxeS 16 'ii'xd8 l:taxd8 17 �xc2 ltJg4 1 8 l:l.d2 i.h6 1 9 l:!.e2 l:!.xd6 with compensation for the pawn, Piket-Kasparov, Wijk aan Zee 2000.

f) White can also advance immediately in the centre: 8 e5 ltJd7. The pattern of play resembles the variation from Game 1 7, except that there is a knight instead of a pawn on a6. It seems that this change does not hurt Black's chances; in lines like 9 e6 ltJb6 10 exf7+ �h8 his king is not the safest, on the other hand his pieces are active.

8 ... c5 9 d5 e6 10 0-0 exd5 11 exd5 (D)

B

We have reached the main-line position of the 7 . . . ltJa6 system. The sole remnant of White's

pawn-centre is on d5. This pawn possesses all the virtues of both a centre pawn and a passed pawn. On the other hand, it lacks the support of its colleagues, which gives it the increased vulnerability typical of an isolated pawn.

1 1 ••• .trs This is the most natural place for the bishop.

Other moves: a) l l . . . b5 is a radical attempt to liquidate

the bone of contention immediately. After 1 2 ltJxb5 'i¥xd5 1 3 i.f4 or 1 2 'ii'xb5 ltJxd5 1 3 i.g5 the weakness of the queenside pawns condemns Black to defence, and so White keeps a modest advantage.

b) l l . . .b6 (D) plans to develop the bishop along its other diagonal, with the hope of coming out at a6. Then:

w

bl) 1 2 l:tdl is met by 1 2 . . . ltJb4, when after 1 3 'ii'h4?! ltJc2 14 l:!.bl i.f5 1 5 i.g5 ltJd4 1 6 l:tbc 1 h6 1 7 i.xh6 ltJxd5 1 8 i.g5 ltJxc3 1 9 ltJxd4 'ii'xd4 i t i s White who has to think about equalizing. Today the search for advantage starts after 1 3 i.g5 h6 14 i.xf6 i.xf6 1 5 l:.d2 or the prophylactic retreat 13 'ii'b3 i.f5 14 ltJel .

b2) 1 2 'ii'h4 has the idea of not wasting time on the d5-pawn's defence and switching at once to the attack on the kingside, using a pin on the d-file and a temporary weakening of the long h l -a8 diagonal as additional tactical trumps. Now 12 ... ltJc7 is the most prudent course. In-stead 12 ... ltJxd5 may be somewhat rash, al-though an attempt to refute the pawn capture immediately doesn't come off: 1 3 ltJxd5 1i'xd5 14 ltJg5 h6 1 5 i.f3 hxg5 16 i.xg5 'iVd4 17 i.xa8 'ii'xh4 1 8 i.xh4 i.e6 19 i.e4 i.xb2 with sufficient compensation. The critical line runs 1 3 i.g5 1i'd6, when after 14 l:tadl ltJac7 1 5 ltJxd5 ltJxd5 1 6 i.c4 i.b7 17 l:td2 f6 1 8 l:!.fdl fxg5 1 9

RUSSIAN SYSTEM: WHITE PLAYS 'iib3 95

lLlxg5 l:tf5 20 'ii'xh7+ �f8 Black not only survived but in the subsequent struggle even managed to win in the game Volkov-Sutovsky, Moscow 2005. 14 lLlxd5 'ii'xd5 15 l:tad l , (FreiSiefring, correspondence game 200 1 ) is probably stronger.

c) Another popular move is l l . . .l:te8 (D), primarily with a view to the struggle for the e4-square.

w

Placing a rook on d 1 is an obvious solution to the problem of supporting the passed pawn, and if the queen's rook is to be entrusted with this task, it is time to bring out the bishop.

c l ) 12 i.f4 doesn't look bad, as the bishop controls a portion of its protege's itinerary, and 12 . . . i.f5 can be met by 1 3 l:tadl . If the thematic 13 . . . lLle4 is countered with the obvious 14 i.d3, Black can equalize by force: 14 . . . i.xc3 1 5 bxc3 b5 16 'ii'xb5 lLlxc3 17 'ii'xa6 i.xd3 1 8 'ii'xd3 lLle2+ 1 9 �hl lLlxf4. Complex play ensues after 1 4 lLlb5, with Black having his share of trumps.

c2) 1 2 i.e3 develops the bishop in more modest fashion. Here it is more secure, attacks a pawn and blocks the e-file. After 12 . . . i.f5 1 3 I:tadl (D) two continuations are employed:

B

c21 ) The first is 1 3 . . . lLle4 14 lLlxe4 l:txe4 1 5 'ii'c 1 , with two threats - a capture on a6, then on c5, and i.d3. After 1 5 . . . 'iib6 1 6 i.d3 the complications favour White, while in regard to 1 5 . . . 'iff6 we shall mention the game Ehlvest-Kriventsov, National Open, Las Vegas 2004: 1 6 i.d3 l:tb4 1 7 i.xf5 'ifxf5 1 8 b3 l:td8, and by sacrificing a pawn with 1 9 lLld2 l:txd5 20 lLlc4, White shut the b4-rook out of the game.

c22) Black's other continuation is 1 3 . . . 'ii'b6, when 14 b3 looks logical. The discussion initially centred on the exchange sacrifice 14 . . . l:txe3 (usually only a temporary one); then with Kasparov's blessing 14 .. Jlad8 became the principal move, which is what the chief expert of today's chess chose: 1 5 lLla4 'ii'd6 16 lLlxc5 lLlxc5 1 7 i.xc5 'it'd? 1 8 i.d4 a6 1 9 i.d3 b5 20 i.xf5 'ii'xf5 2 1 'ii'c6 lLlxd5 22 i.xg7 �xg7 with equality, The Baron-Rybka, Leiden 2007.

c3) There is no unequivocal answer to the question of which rook should stand on d 1 , so many prefer to employ the other rook immediately: 1 2 l:td 1 . If in response to 1 2 . . . i.f5 White develops the bishop to e3 or f4, then in comparison with the positions examined above the only difference is that a different rook is placed on d l , which has no significant importance in the resulting variations. Therefore we shall examine 1 3 d6 instead - the rook went to d l to support the pawn's advance, so the pawn advances. Here 1 3 . . . lLle4? is bad due to 14 d7 l:te7 15 lLlxe4 l:txe4 16 i.g5 i.f6 1 7 i.xf6 'ifxf6 1 8 'ii'c3, when the passed pawn will cost Black dearly. Aside from its career ambitions, the pawn opened the diagonal to the f7-pawn for the queen. If Black allows lLlg5, he has a defence, of course, but it is rather uncomfortable, so prophylactic 13 ... h6 is preferred. The subsequent plan is to commence play on the queenside, where Black has an extra pawn: 14 i.f4 lLld7 15 l:td2 lLlb4 16 'ii'b3 lLlb6.

12 l:tdl Here too, the main alternatives are those

same moves by the dark-squared bishop. We shall start with 1 2 i.e3. The reply 1 2 . . . l:te8 has already been considered (line 'c2' of the previous note); another idea is 1 2 . . . 'ii'b6 (D).

On this square the queen is eyed by the e3-bishop, which gives rise to the idea of preparing b4, introduced by 1 3 lLlh4 i.d7. The knight is


w

poorly placed on h4, but driving the bishop back makes the b 1 -square accessible, which in turn neutralizes the threat of the capture on b2. Likewise, the black bishop has retreated from its active post, but d7 is a useful workplace, from where it supports a possible advance of the queenside pawns. The fact that the enemy queen hampers such an advance provides some consolation to White, and he can set up his threat with 14 a3 or 1 4 l:tabl . However, it transpires that 14 . . . l:tfe8 parries b4, since the exchange sacrifice on e3 with the subsequent capture on b4 allows Black to develop a serious initiative: 14 l:tab1 l:tfe8 1 5 b4 l:txe3 16 fxe3 cxb4 1 7 iff4 ii'c5 1 8 tt:Je4 tt:Jxe4 1 9 ifxe4 l:te8 20 ii'f4 ifxe3+ 2 1 ifxe3 l:txe3 22 ..txa6 bxa6 23 l:txb4 l:td3 with more than sufficient compensation, Timman- Kasparov, Sarajevo 1 999. Naturally, there are more solid continuations, first of all 1 3 b3 l:tfe8 14 l:tad I with complex play and chances for both sides.

If the bishop comes out a square farther by 1 2 ..tf4, we are left with that same 12 . . . ifb6 to examine, when 1 3 ..te5 has original content. By defending b2 indirectly, White neutralizes the mighty bishop on g7 and is ready to build up his pressure in the centre. Accurate defence is required of Black.

12 ... ii'b6 (D) 13 tt:Jh4 Taking advantage of the fact that the queen

has temporarily blocked the way for the bpawn, White decides to drive away the bishop, an idea that we have already seen.

1 3 ii'h4 or 1 3 d6 is more common, but in neither case have White's results been anything to boast about. For example, after 1 3 d6 l:tad8 14 tt:Jb5 (or 14 tt:Ja4 ii'c6 1 5 ..te3 tt:Je4 1 6 ifb5 ..td7, Anand-Kasparov, Frankfurt (rapid) 1 999)

w

1 4 . . . l:tfe8 1 5 tLlg5 l:td7 the raid is over, and now all the pieces will have to go back, LugovoiAreshchenko, Chigorin memorial, St Petersburg 2002.

13 .•• ..td7 14 h3 White combines completing his development

with restraining the black pieces. Before placing the bishop to f4, it is useful to deprive the black knight of the g4-square. However, such unhurried play doesn't pose Black any serious problems.

14 ••. l:tad8 15 ..tr4 .:.res 16 l:td2 ii'a5 (D)

w

Black is fully mobilized and about to produce his arguments, first of all the pawn advantage on the queenside.

17 ..tg5 This is more a defensive, rather than an at

tacking, manoeuvre. Black was threatening to play . . . b5, and the bishop move vacates a decent square on f4. However, this move commits White to surrendering his bishop for a knight.

17 ••. h6 18 ..txf6 ..txf6 19 tt:Jf3 b5 20 iff4 ..txc3

Objectively, this decision is quite correct, but it maximizes the asymmetry of the position and rocks the boat considerably. 20 . . . ..tg7 is

RUSSIAN SYSTEM: WHITE PLAYS 'ifb3 97

much more reliable, and grants Black excellent prospects.

21 bxc3 'ifxc3 22 .:ad1 c4?! (D) After 22 .. .'�g7 23 a4 'ii'b4 24 'ii'xb4 li:Jxb4 25

axb5 Black has the more pleasant position, but there are obvious drawish tendencies.

w

While keeping the full extent of his queenside achievements intact, the text-move gives White an opportunity to activate the knight, after which White's resources - the strong passed pawn and the insecure black king - begin to acquire real dimensions.

23 li:Jd4 'ili'a3? And this is a definite mistake. It is necessary

to bolster the control over the e6-square by 23 . . . lt:Jc5 24 'ifxh6 lt:Je4 25 .:c2 'ifa3, when Black is not worse, at any rate.

24 lt:Je6 (D)

B

It looks like Black didn't calculate to the end the complications started by this move.

24 ... fxe6 24 . . . .U.b8 fails to help. After 25 'iff6 ..txe6 26

dxe6 .:.xe6 27 .:.d8+ .:!.e8 28 .:!. I d7 'iff8 Black is

holding on, but in the line 25 'ifxh6 .i.xe6 26 dxe6 .:.xe6 27 .i.g4 .l:tee8 28 .:!.d6 White's threats cannot all be parried.

25 dxe6 'iff8 26 'ii'g4 ..txe6 26 . . ..l:he6 is probably more resilient. Now in

the ending after 27 'ifxe6+ ..txe6 28 .:!.xd8 the presence of the pawn on g6 surely doesn't hurt Black any, but he is not likely to hold out after 27 .:!.xd7 :deS 28 ..tf3.

27 'ifxg6+ 'ii'g7 28 'ifxg7+ q;xg7 29 .U.xd8 .:!.xd8 30 .:!.xd8 (D)

B

And so, the complications have resulted in an extra exchange in the endgame. The enemy passed pawn demands from White a certain degree of accuracy.

30 .•• lt:Jb4 It was more practical to fish for salvation

chances in lines like 30 . . . c3 3 I .:!.d I li:Jb4 32 ..txb5 c2 33 .:c i ..txa2 34 ..ta4 ..tbl 35 'iii>fl li:Jd3 36 ..txc2 lt:Jxc l 37 ..txb l . The bishop ending that now appears is absolutely hopeless.

31 a3 c3 32 axb4 c2 33 .:!.d1 cxd1'if+ 34 ..txd1 'iii>f6 35 f4 ..tc4 36 'iii>f2 aS 37 bxaS b4 38 'iii>e3 b3 39 'it>d4 b2 40 ..tc2 ..tn 41 g3 .i.xh3 1-0

At the close of this chapter let's recapitulate. In the system with 'ifb3 Black obtains sufficient counterchances; this applies in particular to the continuations 7 . . . a6 and 7 . . . lt:Ja6. White rarely manages to struggle for victory under the comfortable and safe conditions of a small yet solid advantage. More often than not he has to go for a sharp game of calculation, in which both players can reasonably hope to obtain winning chances.

7 Lines with JLf4 or JLg5

As we have seen, the principal methods for White to build a pawn-centre involve a significant degree of strategic risk: either a delay in development or the queen's early entrance into the centre. This is not to everyone's taste, and in this chapter and the next we examine another approach: White postpones the determination of the pawn configuration and first of all solves the question of piece placement. Naturally, the number of reasonable set-ups is fairly large, but e3 is the move common to all of them. We shall divide our coverage of the material into two chapters. The present chapter will focus on lines where the queen's bishop is brought out to f4 or g5 at an early stage, while in Chapter 8 it stays in place for the time being.

Before moving on to specifics of the move .if4 and .ig5, it is worth noting one detail typical of the games in this chapter: the move e3 cuts off the bishop's retreat to its own camp, and Black's counterplay frequently exploits the weakening of the e l -a5 diagonal.

At first glance the post on g5 is the more aggressive of the two, as it creates an indirect attack on the d5-pawn, and, as we already noted, any means of defending this pawn involves positional concessions. However, Black has the counterattacking manoeuvre . . . tt:le4, which is the usual reply to .ig5. The variations where the bishop is brought out at once - 4 .ig5 - are examined in Game 20 (Nakamura-Flu via). If the preliminary 4 tt:lf3 .ig7 is included and only now 5 .ig5, then that is the topic of Game 21 (Kuzubov-Elianov).

The development to f4 looks less concrete. It is true that once the d4-pawn is adequately defended, a real threat will appear, viz. a capture on c7 after taking twice on d5. However, Black usually agrees quite readily to make this a sacrifice, for which he obtains sufficient compensation. There are several ways for White to combine the move .if4 with slightly differing development of the other white pieces, and it is important to be familiar with the subtle but often important differences between them. Game 22 (Nakamura-Kudrin) examines the move-order 4 .if4 .ig7 5 e3, whereas the features unique to 5 tt:lf3 and 5 l:tc 1 are expounded in Game 23 (Lechtynsky-Stocek).

Game 20 [080]

Hikaru Nakamura - Jordi Fluvia Poyatos Barcelona 2007

1 d4 tt:lf6 2 c4 g6 3 tt:lc3 d5 4 .ig5 (D) Attacking the d5-pawn at once. 4 .•. tt:le4 This is the main reaction to the bishop's ap

pearance on g5, and a typical idea in many other cases where a bishop attacks a knight in the opening.

The gambit 4 . . . .ig7 5 .ixf6 ..txf6 is occasionally seen. Now there is no point in the capture with the knight: after 6 tt:lxd5 .ig7 the number of attacks on d4 will always exceed the

B

LINES WITH Ji...j4 OR Ji...g5 99

number of defences, and Black will win the pawn back with a good game. Hence 6 cxd5, when the usual reply is 6 . . . c6. If White accepts the sacrifice at once by 7 dxc6 ..ixd4 8 cxb7 ..ixb7, then Black has decent compensation. The indirect defence 7 l:tc l is more crafty, but only aspires to a modest plus, as the pawn can be recovered in two ways: either 7 . . . 0-0 8 dxc6 ..ixd4 9 c7 ..ixc3+ 10 l:txc3 '5'xdl + 1 1 'it>xd1 lt:Jc6, followed by a methodical rounding-up of the c7-pawn and a slightly inferior ending; or 7 ... '5'b6 8 e3 11i'xb2 9 ..ic4, when White retains his superiority in the centre.

5 ..ih4 The first question that springs to mind is why

not take the knight with 5 lt:Jxe4? This capture occurs occasionally, mainly with a view to creating a fresher position, rather than seeking an objective advantage. White plans to exchange off the e4-pawn and so obtain a more central pawn-mass and meet the freeing . . . c5 with the d5 advance. However, in the meantime his darksquared bishop is placed offside, unlike the master of the long diagonal on g7. A kingside pawn advance implies castling queenside, which is a double-edged idea at best.

5 cxd5 can be dismissed as trickery. Black can switch to play typical of the system examined in the main game with 5 . . . lt:Jxc3 6 bxc3 'ii'xd5, but more interesting is 5 . . . lt:Jxg5 6 h4 lt:Je4 7 lt:Jxe4 'ii'xd5 8 lt:Jc3 '5'a5, when White still has to find a way to demonstrate the benefit of the pawn's position on h4.

Let's consider the bishop's other natural retreat, 5 ..if4 (D).

B

Whereas from h4 the bishop indirectly hinders the advance of the e-pawn with a pin, here it controls another diagonal, takes direct control of

the e5-square and supports operations on the cfile. Following 5 .. . lt:Jxc3 6 bxc3, 6 .. . dxc4 is not as attractive any more, since 7 '5'a4+ and 8 'ii'xc4 places the c7-pawn under attack. On the other hand, the exchange of queens is fine for Black in the line 6 . . . i..g7 7 lt:Jf3 c5 8 e3 0-0 9 cxd5 cxd4 10 cxd4 'ii'xd5 1 1 ..te2 and now 1 1 . . .11i'a5+ 1 2 '5'd2 lt:Jc6 or 1 l ...lt:Jc6 12 0-0 ..if5 1 3 '5'a4 'ii'a5. Then a minor-piece ending is in prospect, with White enjoying a minimal advantage thanks to his slight superiority in the centre, but Black has every reasons to expect to level the game.

5 ... lt:Jxc3 5 . . . c5 (D) has been popular in the recent past.

Black aims to utilize the e l -a5 diagonal as soon as possible, but the idea has lost much of its following.

w

White has two continuations: a) 6 e3 leads to original play. White strength

ens the centre, not fearing the pin 6 . . . '5'a5. His idea is 7 'ii'b3, defending and simultaneously attacking d5 and, given an opportunity, b7 and f7. However, 7 . . . cxd4 8 exd4 ..ih6 (a novel place for the bishop, targeting the d2-square which has suddenly become vulnerable) 9 lld l (the natural 9 lt:Jf3? is bad due to 9 . . . g5 10 ..ig3 g4, when Black smashes though to d2) 9 . . . lt:Jxc3 10 bxc3 i..e6 gives Black good chances. This line, popular some twenty years ago, is a rare guest in current practice.

b) 6 cxd5 leads to a well-known position after 6 . . . lt:Jxc3 7 bxc3 '5'xd5 8 e3 (D).

Now: b1) Black can agree to the queen exchange

by 8 . . . cxd4 9 'ii'xd4 '5'xd4 1 0 cxd4, but then the h4-bishop turns out to be placed well, keeping the e7-pawn in its sights. The pawn must


B

advance, and it is expedient that it should do so at once by IO . . . e6, while Black retains control over the d6-square; with the bishop already fianchettoed the white knight won't be able to resist the temptation to try to occupy that square. The other problem is developing the queenside without letting White infiltrate via the b- and c-files. Even if Black manages to overcome these difficulties and exchange off all the rooks, the resulting endgame has to be held with passive defence, which is, of course, not everyone's cup of tea: I I lt:Jf3 i..e7 1 2 i..xe7 �xe7 13 h4 (aiming to fix the black kingside pawns on light squares) 1 3 . . . h5 1 4 i..d3 lL'lc6 1 5 �d2 i..d7 16 l:tab 1 l:tab8 1 7 e4 l:thc8 1 8 .l:t.hc 1 lt:Jd8 1 9 lt:Je5 l:txc l 20 l:txc l l:tc8 2 1 l:txc8 i..xc8 22 f3, Jobava-Macieja, European Ch, Kusadasi 2006.

b2) Black can play 8 . . . i..g7 first, so that the capture on d4 should not force the exchange of the queens (if Black prefers 8 . . . lt:Jc6 for the same purpose, White can seek another version of the exchange by 9 'ii'f3, as a retreat along the d-file is bad because of 1 0 i..c4, but 9 . . . 'ii'f5 is an option). Now 9 'ii'f3 'ii'd6 1 0 i..b5+ lL'ld7 1 1 lL'le2 cxd4 1 2 exd4 leads to a double-edged game, but if White develops with natural moves, Black's difficulties persist; for example, 9 lL'lf3 cxd4 10 cxd4 lt:Jc6 1 1 i..e2 0-0 1 2 0-0 i..f5 1 3 lt:Jd2 'ii'd7 1 4 i..f3 lt:Ja5 1 5 'ii'c 1 l:tae8 1 6 'ii'a3, Jobava-Brkic, European Ch, Warsaw 2005.

6 bxc3 dxc4 (D) Black's unwillingness to defend passively in

the lines considered above in line 'b' of the previous note has brought this continuation to the forefront. Black voluntarily cedes the centre, but in return hopes to hold on to the extra pawn for a while or else take advantage of the enemy queen's early development.

7 e3

w

Or: a) If White doesn't feel like offering a gam

bit, then 7 'ii'a4+ is the only way and the only time to regain to the pawn. This continuation has not enjoyed much popularity, one of the reasons being the line 7 .. . 'ii'd7 8 'ii'xc4 b6, when the bishop succeeds in getting out to a6, intending an exchange on fl and so depriving White of the right to castle. In recent years White has made several attempts to prove that his potential superiority in the centre counts for more than the loss of castling, but has yet to achieve consistent success: 9 i..g3 c5 10 i..e5 i..a6 1 1 'ii'b3 f6 1 2 i..xb8 l:txb8 1 3 lL'lf3 cxd4 1 4 cxd4 e6 1 5 e4 i..xfl 1 6 �xfl a5 17 h4 a4 1 8 'ii'e3 i..e7, MalakhatkoHarika, Gibraltar 2008. Black strives to generate counterplay based on his passed pawn on the queenside.

b) 7 e4 blocks the long light-squared diagonal and allows 7 . . . b5. This type of formation is playable in some openings, but there Black has usually exchanged his bishop on c3, when White can hope to dominate the dark squares. Here, on the other hand, not only is the f8-bishop alive and well but its counterpart enjoys a rather poor view from h4.

7 .•• i..e6 (D) 7 . . . b5? is clearly bad here: 8 a4 c6 9 axb5

cxb5 1 0 'it'f3. Therefore the pawn has to be defended in an unnatural fashion, although Black's suffering has a good cause.

8 'ifbl Besides this move, 8 lt:Jf3 and 8 l:tbl are also

played. White aims to provoke . . . b6, in order to combine later the weakening of the diagonal with the threat of winning back the c4-pawn and the harassment of the e6-bishop.

After 8 lL'lf3, here is an example typical of both sides' chances: 8 ... i..g7 9 l:tbl b6 1 0 lL'ld2

LINES WITH Ji.j4 OR Ji.g5 101

w

0-0 I 1 .te2 c5 1 2 .tf3 cxd4 13 cxd4?! li:ld7 14 .txa8 'it'xa8 with ample compensation for the exchange, Moiseenko-Svid1er, Russian Team Ch, Sochi 2005. If Black tries to refrain from playing . . . b6, then another typical pattern of play is possible: 8 . . . c6 9 lt:lg5 .td5 10 e4 h6 1 1 exd5 hxg5 1 2 dxc6 lt:lxc6 1 3 .txg5 'it'd5. In this case too, the radical violation of the relative equilibrium promises an uncompromising fight.

In response to 8 .:b1 the reply 8 . . . lt:ld7 is popular, agreeing to return the pawn for the sake of rapid development. Fans of non-standard material balances may be interested in the curious forced line 9 'it'a4 .td5 10 .txc4 .txg2 1 1 'it'b3 .th6 1 2 .txf7 + 'iitf8 1 3 .td5 lt:lc5 14 .txg2 lt:lxb3 15 .:xb3. White has only two pieces for the queen, but the difference in the activity of the opposing armies is evident. After 15 . . . b5 or 15 . . . b6 it is illogical to win the exchange. On the other hand, now the knight has a path to c6 open and if White succeeds in capturing the e7-pawn, his superiority in the centre will be a weighty factor. Nevertheless, Black's extra material should be more important.

8 ... 'it'd5 (D) In reply to 8 . . . b6, White can choose the inter

esting 9 lt:lh3, when 9 . . . .txh3 leaves White's light-squared bishop unopposed and is not prudent. Black is better advised to give up his dark-squared bishop to prevent lt:lf4 and play 9 . . . .th6. After I 0 .tg5 .i.xg5 1 1 lt:lxg5 'ii'd5 the position is simplified, although a small advantage remains for White.

The text-move brings the queen out to d5 not only to defend the b7-pawn but also for a transit stopover on the way to a5.

9 a4 Before this game White usually chose be

tween knight moves here. The f4-square appears

w

attractive but due to the threats of . . . e5 and . . . g5 requires additional security measures. In reply to 9 lt:le2 Black should prudently withdraw with 9 . . . .tc8, protecting the b7-pawn and freeing the queen from guard duty. 9 li:lf3 looks more natural, intending a transfer to d2 once development is complete. Aronian-Svidler, Tal Memorial, Moscow 2006 continued 9 . . . lt:ld7 I 0 .te2 .tf5 1 1 'it'b2 .tg7 1 2 0-0 lt:lb6 1 3 a4 a5 14 li:ld2 .td3. Black didn't really plan to come out from the opening with a healthy extra pawn, of course, and is ready to return it for the sake of completing his development harmoniously.

The text-move aims at making Black's knight lose interest in a trip to b6 and, according to the situation, can lead to an exchange of queens on b5 or a further pawn advance to a5.

9 ••• lt:lc6 Black does not begrudge giving up a couple

of pawns - 10 'it'xb7 .:bs I I 'it'xc7 .th6 - for such an advantage in development.

10 li:lf3 .th6 (D)

w

ll .te2 Noting that the black bishop is no longer a

threat to the a3-square, and a comfortable way to defend the c3-pawn has appeared, White has


been encouraged to advance in the centre in some games: 1 1 e4 'ii'a5 1 2 l:ta3.

ll ... g5 (D)

w

Planning to castle queenside and taking advantage of the fact that the g2-pawn is left unprotected, Black decides to advance against the opponent's likely castled position.

12 .tg3 The complications after 12 .txg5 .txg5 13

tt::lxg5 'ii'xg5 14 'ii'xb7 'oti>d7 are not in White's favour, but 12 e4 'ii'a5 1 3 0-0 deserved consideration.

12 •.. g4 13 tt::lh4 f5 14 0-0 0-0-0 Black can be happy with the outcome of the

opening. He has succeeded in keeping the pawn and laid the foundation for the development of an initiative on the kingside.

15 .td1 White could have played 1 5 a5, in order to

use this pawn to disturb the enemy king, while also freeing the a4-square for the rook. The immediate capture 1 5 . . . tt::lxa5?, perilous in itself, runs into a tactical refutation: 16 .i.e5 l::thf8 17 .txg4 fxg4 1 8 e4. Therefore on the next move the queen assumes blockading duties, uncharacteristic of so respectable a piece.

15 ••• 'ii'a5 16 .tc2 l:thf8 17 'ii'h2 .tg5 18 l:tab1 b6 19 l:tbe1 .txh4

This way Black gets rid of the pressure on f5 and gains an opportunity to activate the other bishop. However, the exchange of such a valuable piece for a knight that risked being left out of the game looks strange. The logical continuation was to continue hindering White's freeing advances; for example, 1 9 . . . .td7 or 1 9 . . . a6, when the only possible active move, 20 f4, renders the g3-bishop passive after 20 . . . .tf6.

20 .txh4 .td5 21 .tg3

White could have chosen 2 1 e4 fxe4 22 .txe4 .txe4 23 l:.xe4 'fi'xa4 24 l:tal 'ii'b3 25 'ii'c I , when the compensation may be sufficient. However, his appetite has grown, and he contrives an ambitious plan of creating a pawn-centre, despite the weakening of his king's position.

21 .•• h5 22 f3 gxf3 23 gxf3 (D)

B

23 ••• f4 This starts a long forced operation that could

have ended in a draw. Black's decision was not forced, as his position is not worse, and he had several attractive continuations at his disposal, such as 23 . . . a6 or 23 . . . -.ti>bS.

24 .txf4 tt::lxd4 25 cxd4 l:tg8+ 26 'oti>f2 'ii'd2+ 27 l:te2 l:tg2+ 28 'oti>xg2 'ii'xe2+ 29 l:tf2 (D)

B

29 ••• l:tg8+?! 29 . . . .txf3+ leads to an instant draw: 30 'oti>h3

.tg4+ 3 1 'oti>g3 h4+ 32 'oti>g2 .tf3+ 33 'oti>h3 .tg4+. The choice of another continuation is not justified logically, given that if he wanted to, White could have forced a repetition of moves there too.

30 .tg3 .txf3+ 31 'oti>h3 .i.g4+ 32 'oti>h4 The return to g2 would have obliged Black to

force a perpetual check. The trip forward looks

LINES WITH iJ..j4 OR iJ..g5 103

very risky, but actually only White has winning chances.

32 •. Jifxe3 33 'ifbS 'ifxd4 34 'ireS 'ifd8 35 l:td2 i..d7 (D)

w

36 :d4?! The exchange sacrifice 36 :xd7 �xd7 37

i..e4 allows White to nullify the threats against his king and obtain a lasting initiative. In a few moves' time, in order to play for a win he has to offer this sacrifice in a less favourable situation.

36 •.. e6+ 37 �xhS i..e8+? After 37 . . . 'ife8+ 38 'it>h4 'ifd8+ the repeti

tion of moves can be avoided only with 39 �h3, when 39 . . . :g5 forces 40 l:.xd7 �xd7. Here the white king feels less secure than in the line from the previous note. Perhaps Black is playing for a win as well?

38 �h6 'ife7 39 l:tf4 i..f7?! This move loses at once, but it is already dif

ficult to suggest a defence. 40 'iff6? This allows Black to force a draw. 40 l:txc4

'iff8+ 41 �h7 wins immediately. 40 ..• Wf8+ 41 �h7 (D)

B

Now that's something - the commander in front of his army. Black's following moves are hard to comment on - perhaps he was in timetrouble.

41. .• �b8?! Here there was a drawing trick: 4 1 . . .:g6 42

i..xg6 i..g8+ 43 �h8 i..f7+. This idea was also possible on the next move, but again he failed to find it.

42 l:tf2 eS? 43 WxeS 1-0

Game 2 1 [09 1 ]

Yuri Kuzubov - Pavel Elianov Moscow 2005

1 d4 lDf6 2 c4 g6 3 lDc3 dS 4 lDf3 i..g7 5 i..gS (D)

s ... lDe4 B This reaction is already familiar to us. The

difference from the development of the bishop to g5 a move earlier is that here it is protected, and White has a wider selection of options to choose from. A retreat to f4 or h4 returns us to the previous game, while the exchange on e4 is clearly worse, but the capture on d5 becomes possible and is the main continuation.

If Black is after more novel positions, he can try 5 . . . dxc4, when 6 e3 i..e6 leaves him quite solid, although if White also wants something

spicier there is an original dish on the menu: 6 e4 c5 7 d5 b5 (D).


w

White has a choice: he can go for the murky 8 e5 b4 9 exf6 exf6 10 'ii'e2+ 'ifi>f8 1 1 �e3 bxc3 12 �xc5+ 'iti>g8 1 3 bxc3 liJd7, or prefer the even sharper 8 d6; for example, how do you like this line: 8 . . . �b7 9 dxe7 'ii'xe7 10 e5 liJbd7 ( 10 . . . �xf3 I I 'ii'xf3 'ii'xe5+ 12 �e3 b4 is stronger) 1 1 'ii'e2 �xf3 1 2 exf6.

6 cxd5 ltJxgS Also seen is 6 . . . ltJxc3 with familiar motifs,

except that White hasn't spent a tempo on retreating the bishop and developed the knight instead. So generally speaking, it is seen as more natural and logical to capture the bishop.

7 ltJxg5 e6 If the other pawn steps forward by 7 . . . c6,

then besides accepting the gambit with 8 dxc6 lDxc6 9 d5 or 9 e3 e5 10 d5 'ii'xg5 1 1 dxc6 0-0 1 2 h4 White can play the simple 8 liJf3, and Black's position is more boring than in the main lines. Black should also be ready to meet the tricky move 8 'ii'b3.

8 'ii'd2 8 liJf3 exd5 (D) leads to a quiet game.

w

This position features the so-called Carlsbad set-up, well known from the Queen's Gambit. There are some differences - a different set of

minor pieces, the fianchettoed bishop - but that does not change the cardinal methods of playing such positions. White's principal plan is still the minority attack, i.e. a queenside advance to create weaknesses in Black's pawnstructure there. Once Black has played or been forced to play . . . c6, White will seek to push the b-pawn to b5, when Black will have to decide what type of weakness to choose.

Black has several ways to impede White's progress: he can play the a-pawn to a5, so that in reply to b4 he can exchange or push forward and with the counter-thrust . . . b5 create mutual weaknesses - c6 for himself, b4 (a3) for White, intending to shield his weakness with the transfer of the knight via b6 to c4. Another method is to play . . . a6, and when White carries out a4 and b4-b5, push the pawn further by . . . a5, partially avoiding the opening of lines; Black thus acquires a strong post on b4 and denies the white knight the possibility of manoeuvring via a4. Also, the position of the bishop on g7 sometimes makes the counter-thrust . . . c5 possible in reply to b5.

Of course, Black's main counterplay should be based on an attack on the kingside, although if it includes the . . . f5-f4 advance, then care must be taken that White cannot advantageously block the f5-pawn with f4, followed by occupation of e5 .

In this system the play doesn't follow a forced course, and the players have considerable freedom in determining the desired set-up and the move-order used to reach it. The following serves as an example of some typical themes: 9 e3 0-0 10 �e2 c6 I I 0-0 'l!i'e7 1 2 a3 �e6 1 3 b4 liJd7 14 lDel a5 1 5 liJd3 b5 16 'ii'd2 a4 1 7 lDc5 liJb6 1 8 l:bel .l:f.ae8 1 9 �d3 �c8, Nakamura-V.Mikhalevski, Miami 2007.

We now return to 8 'ii'd2 (D). This queen move leads to more original posi

tions. 8 ... h6 This preliminary attack on the knight side

steps the following sharp and unclear line: 8 . . . exd5 9 'l!i'e3+ 'iti>f8 10 'ii'f4 ..tf6 1 1 h4 h6 (it is a bad idea to tolerate the knight any longer: after I I . . .'iti>g7?! 12 e4 dxe4 1 3 ..tc4 White seizes the initiative), when White can try the speculative 1 2 ltJxd5 i.xg5 1 3 'ii'e5 .l:f.h7 ( 1 3 .. .f6 14 'ii'xc7 'ii'xd5 1 5 'ii'xc8+ 'iti>g7 ! 1 6 'ii'c7+ lDd7 17

LINES WITH iJ..j4 OR iJ..g5 105

B

hxg5 fxg5 gives Black good compensation) 1 4 hxg5 lt:Jc6 1 5 �e4 iJ...f5 1 6 �f3 lt:Jxd4 1 7 �a3+ �g7 1 8 lt:Je3. The simple retreat 1 2 lt:Jf3 is also possible, with a complex struggle.

9 lt:Jh3 The choice of retreat is determined by where

White likes the knight better - on f3 or f4. The Carlsbad formation is most likely here as well, and the post on f4 promises fresher play. The loss of a tempo is not significant in this case; the f3-square is also only a temporary stop before subsequent manoeuvres.

9 ... exd5 10 �e3+ This check vanished from practice after the

current game. Depriving Black of the right to castle is a tempting idea, but the detriment to White's development proves no less significant. Today 1 0 lt:Jf4 is preferred.

10 ... �f8 11 lt:Jf4 (D)

B

ll ... c5 The most principled continuation. In case of

l l . . .c6 1 2 'ii'f3 �g8 1 3 e3 the white pieces are posted well, and the weakening of Black's kingside pawn-formation justifies preparations for an advance on that part of the board.

12 dxc5

1 2 �f3?! is quite suspect, as the threat of taking on g6 does not frighten Black. After 1 2 . . . cxd4 1 3 lt:Jcxd5 �g8 the white king has more problems.

12 ... d4 13 �d2 (D)

B

13 •.• lt:Jc6! This move had not previously enjoyed much

popularity before Elianov demonstrated its power. The knight clearly has two other options:

a) 13 . . . lt:Ja6 plans different routes for the knight, with the elimination of the c5-pawn in the process, but doesn't threaten a capture on c3 and gives White a tempo for an attack against the d4-pawn. The following entertaining draw is a good example: 1 4 e3 ..if5 1 5 .l:!.d 1 dxc3 1 6 �xd8+ .l:!.xd8 1 7 .l:!.xd8+ �e7 1 8 .l:!.xh8 cxb2 1 9 lt:Jd5+ �e6 20 lt:Jf4+ �f6 2 1 lt:Jd5+ �e6 1h-1h Markus-Elianov, European Under- 1 8 Team Ch, Balatonlelle 2000.

b) 13 . . . lt:Jd7 has the positive aspect of forcing the c3-knight to move, but obstructs the bishop. We should warn the knight against taking the path leading to the left: 14 lt:Ja4? ! b5 1 5 cxb6 axb6 1 6 �b4+ �g8, when in FreislerHenrichs, Pardubice 2007 White decided that he could afford the developing 1 7 g3?. In reply Black ingeniously cleared the bishop's diagonal by 1 7 . . . ..if8 1 8 �xd4 ..ig7 1 9 �e4 !txa4 20 �xa4 ..ixb2, and it transpired that White had to give up the rook, remaining a piece down.

14 lt:Je4 g5 (D) For the moment such an advance cannot be

called a weakening. It is hard to make an omelette without breaking eggs, so Black pushes away the white pieces, seizing the centre and ignoring the holes that appear. Now ifthe initiative fails to bring the desired results, then White


w

will be able to raise the issue of weaknesses. For the time being there is no reason for the knight to go to h5 - the exchange of the bishop and the simplification of the position are not a bad idea, of course, but in the process the enemy rooks are connected and the extra pawn has to be returned.

15 tiJd3 1!i'd5 16 tiJd6 ..ie6 (D)

w

The critical position of the game and the entire variation with 1 3 . . . tbc6. White has an extra pawn and great difficulties in completing his development. Another worry is the outwardly beautiful knight on d6 whose position is in fact unstable, and it is hard to see how it could be manoeuvred to a more secure post. White bases his hopes for a successful outcome on the fact that Black is also yet to liberate the h8-rook from imprisonment. At the moment there are no grounds for panic.

17 l:tg1? The capture 1 7 tbxb7 is met by 17 . . . tbe5,

when the second pawn has to be returned, and meanwhile the position has been opened and the black bishops have become more active.

1 7 e4 is a radical attempt to resolve the situation. After 1 7 . . . dxe3 1 8 fxe3 b6 1 9 tiJb5 l:td8 20

tbc7 or 20 tbc3 it is hardly prudent to seek an advantage, but returning the pawn at the right moment gives real chances of equality.

If White has nerves of steel that allow him to tolerate the static onlookers on fl and h 1 , the rest of the pieces can embark on a local operation on the queenside, initiated with 1 7 b4, 1 7 l:tc 1 or 1 7 tiJb5, striving to expel the black forces from the centre and complete development.

The move chosen by White takes the last chance of a trip to the kingside away from the white king. While it is in the centre, getting through to it is not a difficult task.

17 ••• tbe5 (D)

w

Removing the c5-pawn's defender and clarifying the situation in the centre.

18 g3? A pointless loss of a tempo in an already

grave position. 1 8 l:tc I or 1 8 g4 at least made some sense.

18 ... tiJxd3+ 19 exd3 11i'xc5 20 tbe4 1We7 (D)

w

Additionally it transpires that the knight has nowhere to move to.

21 11i'e2?!

LINES WITH iJ..j4 OR iJ..gS 107

Some sort of resistance could have been organized with 2 1 g4, clinging to the e4 post.

21. •• f5 22 tt:'ld2 l:le8 23 0-0-0 'ii'c7+ 24 �b1 'ii'f7

Wishing to extract the maximum. Besides the capture on a2, 25 . . . .id5 is threatened, and there is not even a hint of compensation for the queen.

0-1

Game 22 [082]

Hikaru Nakamura - Sergey Kudrin USA Ch, San Diego 2004

1 d4 tt:'lf6 2 c4 g6 3 tt:'lc3 d5 4 .if4 (D)

B

Before playing e3, White brings out the bishop to a solid central post.

4 ••. .ig7 5 e3 White has several possible plans of develop

ment, and some choice in terms of the moveorder too. One of the key factors is the assessment of the possible . . . tt:'lh5. A number of systems feature the move l:tc l , where the rook hinders . . . c5. Here the text-move and 5 tt:'lf3 appear to be the most natural moves, and with these continuations the black knight's move to the edge of the board fails to yield benefits and looks absurd. The immediate 5 l:tc I is also played, but it is precisely here that 5 . . . tt:'lh5 is not bad. The bishop can return home with 6 .id2 or stay out in the field by 6 .ig5 h6 7 .ih4, reckoning that the tempo it has lost will be outweighed by the time spent on the enemy knight's leap. In both cases there are two plans of approximately equal value: the standard play with . . . dxc4, followed by . . . .ie6, or the the-matic . . . c5 counterblow in the centre. Here is a possible course of events: 6 .ig5 h6 7 .ih4 c5 8 e3 cxd4 9 tt:'lxd5 tt:'lc6 10 exd4 .ixd4 I I 'ii'd2 g5 12 .ie2 tt:'lf6 1 3 tt:'lf3 .ib6 14 tt::lxf6+ exf6 15 'ii'xd8+ .ixd8 1 6 .ig3 .ia5+ 1 7 �fl , and the

outcome of the opening suits Black just fine, I.Sokolov-Svidler, Wijk aan Zee 2004.

We shall examine the variations that are tied to 5 tt:'lf3 and have independent significance in the next game.

5 ••• c5 With White having protected the d4-pawn,

capturing on c7 becomes a genuine threat. If that does not frighten Black, he can choose 5 . . . 0-0 (D).

w

a) The most critical reply is clearly to accept: 6 cxd5 tt:'lxd5 7 ltJxd5 'ii'xd5 8 .ixc7. Then the sacrificed pawn can be regained at once with 8 . . . tt:'la6 9 .ixa6 (a retreat by the bishop not only means a loss of time but also makes the aS-square available to Black, who obtains a sufficient initiative on the queenside) 9 . . . 'ii'xg2, when the endgame after I 0 'iff3 'ifxf3 1 1 tt:'lxf3 bxa6 is considered to give White a small advantage, but is quite acceptable for Black; for example, 12 l:tcl .ib7 13 'it>e2 l:tac8 14 l:thdl f6 1 5 d5 f5 16 b4 .ib2 17 l:tc5 a5. Now the precipitate 1 8 .ixa5? .ic3 ! suddenly presented White with an unpleasant discovery in Vallejo Pons-V.Mikhalevski, Calvia 2007: Black's threats of l 9 . . . l:txc5 and l 9 ... .ia6+ cannot be parried simultaneously. Black also has a riskier but more belligerent


option: 8 . . . ltJc6 9 lDe2 (the c3-square is more enticing than f3) 9 . . . .i.g4 10 f3 .l:.ac8 I I lDc3 'ii'e6 1 2 .tf4, and here Black sacrifices a piece with a capture on d4, obtaining a strong attack, although it is not clear whether it is sufficient.

b) Naturally, White is not compelled to take on c7. Let's dwell a bit on the move 6 'ii'b3 (D).

B

The idea behind this lunge is the same as we saw in Chapter 6 - to put pressure on d5. Ceding the centre by 6 . . . dxc4 is less appealing than in the Russian System, as White is ready to recapture with the bishop, not allowing Black the fun of chasing the queen about. The choice of the reply comes down to this: should Black defend or counterattack? Defence is 6 . . . c6. This position often arises from the Slav Defence and can equally well claim to belong to that opening. Black is not likely to manage to maintain his grip on d5, but for now he opens a road to a5 and lays down preparatory work for the deploy-ment of the queenside by . . . dxc4, . . . b5, . . . i.e6 and at the right moment . . . b4; he will also at-tempt, one way or another, to avoid being left with a weak pawn on c6.

The counterattack 6 . . . c5 also shares the field with another opening, this time the Caro-Kann Defence. Events can develop as follows: 7 cxd5 cxd4 8 exd4 tiJbd7 9 i.e2 tiJb6 I 0 .i.f3 .i.f5 I I l:tdl 'ii'd7 1 2 h3 h5 1 3 lDge2 .l:.fd8 14 d6 exd6 with chances for both sides.

c) The main difference between the various lines is in the order in which the moves e3, lDf3 and .l:.c I are combined. In this game White began with e3 ; let's examine it in combination with 6 .l:.c I . Here too, 6 . . . c6 and 6 . . . dxc4 are played, but 6 . . . c5 must be recognized as the most principled continuation, and we shall pause to examine it in some detail. After 7 cxd5 cxd4

8 'ii'xd4 lDxd5 there is no question of an advantage, so White chooses 7 dxc5, to which 7 . . . i.e6 (D) is considered to be the most solid reaction (another idea is 7 . . . 'ii'a5 8 cxd5, though Black hasn't demonstrated full compensation with either 8 . . . .l:.d8 or 8 . . . lDe4).

w

Now 8 ltJge2 is sometimes played, intending after a double exchange on d5 to occupy the vacant c3-square, but the obvious and principal continuation is 8 tiJf3 - a normal developing move that also offers prospects of disturbing the e6-bishop from d4 or g5. We shall note these replies to the natural continuation 8 . . . lDc6 (D):

w

c I ) In case of 9 tiJd4 liJxd4 I 0 exd4 dxc4 Black has nothing to complain about, with his control over d5 and a subsequent undermining . . . b6.

c2) After 9 ltJg5 Black must choose between 9 . . . 'ii'a5 10 .i.e2 :adS I I tDxe6 fxe6 1 2 'ii'a4 'ii'xc5 and the razor-sharp line with a piece sacrifice 9 . . . i.g4 10 f3 e5 I I cxd5 exf4 1 2 dxc6 'ii'e7.

c3) 9 .te2 lDe4 leads to a quieter game; for example, 10 0-0 tDxc3 1 1. bxc3 dxc4 1 2 ltJg5

LINES WITH JLJ4 OR JLg5 109

.td7 13 .txc4 lt::la5, when the condition of the extra pawn leaves much to be desired, while Black has a normal position. w

d) Let's move on to another move-set - e3 + lt::lf3, that is 6 lt::lf3 (D).

B

Here as well the most interesting continuation is 6 . . . c5. For the same reason as in line 'c' White replies with 7 dxc5, and Black tries to take advantage of one of the newly-opened diagonals using the already-familiar methods:

d l ) 7 . . . lt::le4 - the bishop's desire to find work on the long diagonal is commendable. White's attempt to neutralize the pressure with his own bishop promises little: 8 .te5 .txe5 9 lt::lxe5 lt::lxc3 10 bxc3 'ika5. After 8 l:.cl lt::lxc3 9 bxc3 there is a choice between the acceptable ending 9 . . . dxc4 10 'ikxd8 .l:txd8 1 1 .txc4, with real chances of regaining the pawn, and 9 .. . 'ika5, when Black is ready for a different version of the pawn sacrifice: I 0 cxd5 lt::ld7 I I 'ikb3 lt::lxc5 1 2 1i'b4 1i'b6, with compensation.

d2) 7 . . . 1Wa5 is another well-trodden path. Black threatens to include two more combat units at once in the attack on c3, so it's time to take defensive measures: 8 .l:tc I dxc4 (8 . . . lt::le4 9 cxd5 lt::lxc3 I 0 bxc3 transposes to a line given just above in line 'dl ') 9 .txc4 1i'xc5 (D).

The pawn-centre has disappeared, leaving a symmetrical formation. Often this translates into the reduction ofthe positional tension, but Black still has to neutralize White's small development advantage and the threats against the black queen, whose early entry into the centre has left it facing an assault by the enemy pieces. The immediate attempt to take advantage of this circumstance with 10 lt::lb5 is parried by 10 . . . .te6, when 1 1 .txe6 runs into 1 1 . . .'ihb5, 1 1 .td3 (or 1 1 .te2) is met with

I l . . .'ikh4+ 1 2 1i'd2 'ikxd2+ and 1 3 . . . lt::la6, while grabbing the exchange by I I lt::lc7 is risky due to l l . . ..txc4 1 2 b3 'ikh4+ 13 'ii'd2 'ii'xd2+ 14 lt::lxd2 .td3 1 5 lt::lxa8 lt::ld5, Inkiov-Lputian, Saint John 1988. Therefore White completes his development with 10 .tb3 1i'a5 1 1 0-0 lt::lc6, retaining superior piece control over the centre squares. The main events develop after 1 2 h3 .tf5 1 3 'ii'e2 lt::le4 14 lt::ld5 e5 : the sacrifice of the exchange for two pawns by 15 .l:txc6 bxc6 16 lt::le7+ 'iii>h8 17 lt::lxc6 1i'b6 1 8 lt::lcxe5 .te6 is not considered dangerous, although after 1 5 .th2 Black still experiences some difficulties.

By choosing an earlier . . . c5 (as in our main game), Black avoids some of the lines we have examined in this note, including the one with the capture on c7, but the uncastled position of his king grants White a selection of other continuations.

6 dxc5 1i'a5 (D)

w

7 .l:tcl One of the new ideas is 7 'ii'a4+, but this is

not considered to be dangerous for Black. The knight's trip to the edge of the board following 7 . . . 'ifxa4 8 lt::lxa4 allows White's extra pawn to be kept for a while, but as the mobilization of


the black anny progresses, this task becomes increasingly untenable. Here is an example: 8 ... .i.d7 (another reliable continuation is 8 ... 0-0) 9 ltJc3 ltJe4 1 0 ltJxd5 ltJa6 1 1 f3 ltJexc5 1 2 0-0-0 e6 1 3 ltJc3 .i.xc3 14 bxc3 f6 15 g4 0-0-0, Portisch-Ruck, Zalaegerszeg 2004.

7 .•. ltJe4 Black can play 7 . . . dxc4 8 .i.xc4 0-0, when 9

ltJf3 transposes into note 'd2' to Black's 5th move, but White has retained the option of developing the king's knight to e2, where it possesses different functional capabilities: 9 ltJe2 'ii'xc5 10 'ii'b3 ltJc6 1 1 ltJb5 'ii'h5 1 2 ltJg3 'ii'h4 1 3 ltJc7. Now Black is considered to have sufficient counterplay after 1 3 . . . e5, although this whole line is quite sharp and concrete; some preparation is required for players who wish to venture into it.

8 cxd5 ltJxc3 9 'ii'd2 'ii'xa2 10 l:[xcJ This is not so much an exchange sacrifice as

the desire to keep the c5-pawn, with which White has to part after 10 bxc3 (D).

B

White obviously has the upper hand in the centre, but Black has a passed pawn ready to advance, which, supported by the g7-bishop, can quickly become a real threat. The immediate queen exchange 10 . . . 'ii'xd2+ 1 1 'it>xd2 has been tested, as without the queens an advance in the centre does not look as fearsome, but then in order to regain the pawn Black has to acquiesce to other concessions: 1 1 . . .ltJd7 1 2 .i.b5 0-0 1 3 .i.xd7 .i.xd7 14 e4. White is preparing a secure defence of the extra pawn from e3, and the attempt at counterplay with 14 . . . f5 is met by 1 5 e5, aiming to wall up the fianchettoed bishop; e.g., 15 ... :fc8 16 c4 :xeS 17 .i.e3 and 1 8 f4. The struggle to gain some breathing space for this bishop by 1 5 . . . e6 leads to curtailing the

freedom of the light-squared one: 16 c4 :fc8 1 7 c6 bxc6 1 8 d6 g5 19 .i.xg5 .i.xe5 20 c5.

More often Black retreats with 10 .. .'ifa5, not putting off the elimination of the c5-pawn for long. The main discussion is centred on 1 1 .i.c4 (the bishop takes up an insecure position but is ready to retreat to a2, where it blocks the passed pawn's way and hinders the e7-pawn's advance) 1 1 . ..ltJd7 1 2 ltJf3 ltJxc5 with chances for both sides.

We now return to 1 0 .l:tc3 (D):

B

10 •.• 0-0 Black has to be an inveterate optimist to grab

such an exchange. After 10 . . . .i.xc3 1 1 'ii'xc3 f6 (or first 1 1 . . .'ii'al+ 1 2 'it>d2 f6) White has in compensation a pawn, a powerful centre, an advantage in development and the chronically insecure black king, so we can evaluate this continuation unequivocally in White's favour.

11 .i.c4 'ii'a1+ An attempt to regain the c5-pawn anyway

has found no followers: 1 1 . . . 'ii'a5 1 2 :c2 'it' xeS 13 d6, and White seizes the initiative.

12 :ct 'ii'xb2 13 e4 (D)

B

13 .•• 'ii'a3

LINES WITH il..j4 OR iL.g5 111

In a later game Lukacs-Ruck, Hungarian Team Ch 2005/6, Black preferred 1 3 . . . tLld7 14 ..ie3 and only now 14 . . . 1i'a3 1 5 ..tb5 tLle5 1 6 ..id4 tiJd3+ 1 7 ..ixd3 ..ixd4 1 8 tt:Jf3 ..tg7 with an approximately even game.

14 tiJf3 (D)

B

White's centralized army looks formidable and demands of Black concrete counter-action; merely advancing the a-pawn will not suffice.

14 ... 1i'xc5?! A very risky decision; either way of develop

ing the knight was to be preferred. The forced sequence that follows renders the black king's position critical.

15 d6 1i'a3 16 ..txf7+ '>t>h8 17 ..ig5?! Of several tempting continuations, White

picks one of the least effective. 17 tLlg5 (or 17 h4 at once with an extremely strong attack) 17 . . . tt:Jc6 1 8 h4 looks very menacing.

17 ... tt:Jc6 18 dxe7 tt:Jxe7 19 .l:.c7 1i'a1+ 20 l:lcl 1i'a3 21 llc7 1i'a1 + 22 '>t>e2?

Recklessness. The attempt to take advantage of the brief removal of the queen from the game fails, and from now on it is White who has to seek a draw.

22 •.. 1i'xh1 23 i.xe7 ..ig4

The tempting 23 ... b6? ! can be met by 24 ..ic4 ..ig4 25 ..ixf8 .l:.xf8, when White can at the very least start the pursuit of the enemy rook along the seventh rank by 26 l:lf7.

24 ..txf8 .l:.xf8 25 h3 ..ixf3+ 26 gxf3 1i'xh3 27 .l:.xb7 1i'c8 28 .idS 1i'c5 29 e5?

A sharp move, not justified by necessity. The following part of the game may have taken place in time-trouble. It is doubtful that White was still striving for victory; perhaps he feared the prospect of protracted defence in the endgame, so he attempted to activate his main trump. Actually, after an exchange of queens Black has a purely symbolic advantage at best, so something like 29 1i'b4 is appropriate.

29 •.. ..ixe5 30 1i'h6 ..ic7 31 'iti>f1 (D) 3 1 .l:.xc7? fails because of the zwischenzug

3 1 . . .l:le8+, which is why White withdraws his king.

B

31 ... .l:.b8? Missing the win that could have been se

cured by 3 1 . . . l:lf4, disrupting White's control of the c l -square and threatening 32 . . . 1i'c1 + and 33 .. J:tg4/e4+ or 32 . . . 1i'xd5 with the subsequent route . . . 'ii'd l + - ... 1i'xf3+.

32 1i'h4 1i'cl+ 33 'iti>g2 1i'f4 34 1i'xf4 1/z-1/z

Game 23 [092] v

J iri Lechtynsky - J iri Stocek Czech Ch, Karlovy Vary 2005

1 d4 lLlf6 2 lLlf3 g6 3 c4 ..tg7 4 tLlc3 d5 5 ..if4 (D)

In this game we move on to lines where White first of all plays ..if4 and tt:Jf3 before

deciding on further development. Naturally, the moves e3 and .l:.c 1 are still very much on the agenda.

5 ... 0-0


B

By moving to f3, the knight defended d4, so as in the previous game, the c7-pawn has come under attack. And likewise, Black can parry this attack with the standard undermining thrust 5 . . . c5 prior to castling. This has specific features of its own that distinguish it from the lines we saw in Game 22, the main one being that the f4-bishop's retreat back to the defence is not cut off, and so after 6 dxc5 "ii'a5 7 cxd5 lt:Jxd5 (after 7 . . . lt:Je4 8 .i.e5 or 8 .i.d2 it is difficult to regain the pawn) 8 "ii'xd5 .i.xc3+, White can play 9 .i.d2. Now the line 9 . . . .i.xd2+ 10 "ii'xd2 ..Wxc5 1 1 l:r.c I "ii'f5 1 2 lLld4 "ii'd7 13 "ii'h6 fails to rouse Black's enthusiasm, and Tseshkovsky devised the gambit 9 . . . .i.e6 (D).

w

There are few grounds for seeking an advantage after I 0 .i.xc3 "ii'xc3+ I I "ii'd2, which is well met by both l l ... "ii'xd2+, with 12 ... lt:Ja6 to follow, and 1 1 . . . "ii'xc5 at once. Thus the sacrifice has to be accepted with 10 "ii'xb7 .i.xd2+ I I lt:Jxd2 0-0. Now there is no point in taking the rook at once with 1 2 "ii'xa8, as the reply 12 . . . l:r.d8 threatens 13 . . . .i.d5, and the jab 1 3 b4 allows a perpetual check after 1 3 . . . 'i¥xb4 14 l:r.d1 l:txd2. Therefore 1 2 b4 is played earlier, when the queen is still on b7. Black has two replies:

a) 1 2 . . . "ii'a4 retains control over the c6-square. Here the capture of the rook is again disadvantageous: 1 3 "ii'xa8? lt:Jc6 14 "ii'b7 :b8 1 5 "ii'c7 l:tc8 1 6 "ii'g3 lLlxb4 1 7 l:r.cl lt:Jc2+ 1 8 .l:.xc2 "ii'xc2, and Black has enough activity to win back the material. However, the main continuation 1 3 e3 presents Black with serious problems; for example, 1 3 . . . l:r.d8 1 4 a3 "ii'c2 1 5 l:r.d1 .i.d5 16 "ii'xe7 lt:Jc6 17 "ii'f6 .i.b3 1 8 .i.e2 .i.a4 1 9 b5 lt:Ja5 20 lLlb3, and White succeeded in extricating himself and retaining the material advantage in Akhmetov-Turov, Moscow 2005.

b) 1 2 . . . "ii'a3 moves the queen one square further and intends to counter 1 3 "ii'xa8 with 1 3 . . . l:r.d8. Besides the threat of 1 4 . . . .i.d5 Black keeps in reserve 14 . . . "ii'b2 1 5 .l:.dl l:r.xd2 with a perpetual check on b I and b4 or c I and c3, but it still has to be demonstrated how Black is to equalize after 14 e4, which parries both threats.

The general conclusion is that with this move-order the early . . . c5 is not seeing its best days. But chess, like nature, appears to preserve a balance . . .

6 l:tcl . . . because, on the other hand, there are fewer

reasons to fear the capture on c7 than in the previous game. In the position after 6 cxd5 lt:Jxd5 7 lt:Jxd5 "ii'xd5 8 1L.xc7 the white knight can no longer block the c-file, and Black develops according to the pattern . . . lt:Jc6, . . . .i.f5 and . . . .l:.c8, obtaining prospects of invading on c2. If White disrupts the black knight's planned trip with a3, then Black secures a decent game by taking advantage of the weakened light squares on the queenside combined with the undermining thrust ... e5.

Since 6 e3 returns us to Game 22 (note 'd' to Black's 5th move, to be precise), only the textmove has independent significance.

6 ..• dxc4 (D) The thematic 6 . . . c5 usually transposes after

7 dxc5 .i.e6 into familiar positions. In the lines with the development of the queen to aS White acquires the additional option of opposing it with the bishop on d2.

In taking on c4, Black makes a concession in the centre, but intends to inconvenience White with either threats to retain the c4-pawn or by a counterattack against the centre pawns if White plays e4.

7 e4

LINES WITH i..J4 OR JL.g5 113

w

The alternative is the restrained 7 e3 . Then: a) Black can count on equality by playing

7 . . . c5. In the resulting symmetrical position he must carefully neutralize the activity of White's better-centralized army. After 8 dxc5 'i¥a5 the play boils down to the position examine in note 'd2' to Black's 5th move in Game 22. If White chooses another plan of development by 8 .i.xc4 cxd4 9 tt'lxd4, there are several replies, including 9 . . . tt'lbd7, 9 . . . a6 and 9 . . . 'i¥a5. As an example one may suggest the following set-up: 9 . . . .i.d7 lO tt'lb3 'ii'b6 I I 0-0 tt'la6, LautierSvidler, Biel 200 1 .

b) 7 . . . .i.e6 (D) i s more ambitious.

w

We saw a similar idea in Game 20. White needs to react in concrete fashion: 8 tt'lg5 .i.dS (8 . . . .ig4 also has been tried, aiming to provoke 9 f3 then retreat with 9 . . . .tc8; after preparing . . . tt'ld5 with . . . e6 or . . . c6, Black endeavours to expose the weakness of the e3-pawn) 9 e4 (there is also an amusing continuation with a sacrifice of a piece: 9 tt'lxd5 tt'lxd5 1 0 .ixc4 tt'lxf4 1 1 'ii'f3 - no refutation is apparent, but Black has no trouble equalizing; for example, 1 1 . .. e6 1 2 'ii'xf4 'i¥f6) 9 . . . h6 l O exd5 hxg5 I I .ixgS tt'lxd5 (the sharp 1 l . . . b5 has been tested as well) 12

.ixc4 tt'lb6. The subsequent play must determine which factor is of more importance: the weakening of the black king' s cover or the vulnerability of the white pawn on d4.

7 ••• .tg4 For several years, studies focused on the

sharp 7 . . . b5, but nowadays Black's compensation for the pawn after 8 tt'lxbS tt'lxe4 9 .ixc7 'i¥d7 l O .ixc4 a6 I I tt'la3 tt'lc6 I 2 0-0 .ib7 13 .ib6 is considered insufficient, and the line has disappeared from practice.

7 . . . c5 is possible too. Here 8 . . . b5 is a good reply to 8 dS, so White prefers 8 dxc5 'iVaS, and continues the attack in the centre with 9 eS (D).

B

Black has three replies: a) 9 . . . tt'lfd7 is the most passive. White's

domination in the centre permits him to embark on favourable complications: I 0 .ixc4 tt'lxc5 1 1 0-0 tt'lc6 12 tt'ld5 .ie6 1 3 b4.

b) 9 ... tt'lh5 lO .ie3 is double-edged. The knight has moved to the edge of the board, but Black wins back the pawn on e5, and it is not clear whether White's superiority on the queenside will prove to be the more significant factor.

c) 9 . . . .l:f.d8 lO .id2 tt'lg4 is the sharpest reply. 1 1 .i.xc4 and then:

c l ) I I . ..'iVxcS?! 1 2 tt'le4 'i¥b6 leads to complications that favour White after 1 3 .ixf7+ �xf7 14 .l:f.xc8 .l:f.xc8 15 tt'lfg5+ �g8 16 'i¥xg4. Then 1 6 . . . tt'la6?? loses to 1 7 tt'ld6! , but neither 16 . . . .l:f.f8 17 'ir'h3 nor 1 6 .. .'ii'c6 1 7 0-0! (e.g., 17 ... tt'ld7 I8 tt'ld6 or 17 ... .txe5 1 8 ii'f3 .U.f8 1 9 'ir'b3+ �h8 20 .U.e I) can satisfy Black either.

c2) After 1 l . . .tt'lc6, the lunge 1 2 tt'lbS brings White no benefit because of the exchange sacrifice 1 2 .. Jhd2 13 'ii'xd2 'i¥xd2+ 14 tt'lxd2 tt'lgxe5. White usually has to ret�m the extra material to neutralize Black's activity.


8 .txc4 (D)

B

8 ... l'Llh5 The knight goes to the rim but hopes to take

part in the struggle for the f4-square. The immediate 8 . . . .txf3 allows the 9 'ifxf3 recapture, when Black has two continuations available:

a) He can plant a knight in the centre by 9 . . . l'Llc6 10 d5 t'Lld4 1 1 'ifd3 c5, to be followed by an advance of the queenside pawns.

b) 9 . . . l'Llh5 offers White a choice between parting with his light-squared bishop by 1 0 .tg5 l'Llc6 1 1 d5 l'Lle5 1 2 'ife2 or with the darksquared one with 10 d5. White manages to keep the centre in decent shape, but the resulting exchanges reduce its value.

9 .te3 .txf3 10 gxf3 Here in reply to 10 'ifxf3, 10 . . . .txd4 is quite

playable. 10 ••• e5 Taking into account his opponent's bishop

pair, Black can give the game a more closed aspect by 10 . . . e6. However, after 1 1 f4 'ifh4 (White should meet l l . . .t'Llc6, attempting to shuffle the move-order, with 1 2 d5, but not 1 2 e5 .th6, when White has to reply with 13 l'Lle2, depriving his queen of the post on f3) 1 2 'iff3 l'Llc6 1 3 l'Lle2 l:tad8 14 l:td l , White retains the centre and the better prospects.

11 dxe5 'ifh4? The position after 1 I . . ..txe5 1 2 'ifxd8 l:txd8

is considered fully acceptable for Black; White's advantage of the bishop-pair is offset by his slightly spoiled pawn-structure. White can try 1 2 .th6, when 1 2 . . . .tg7?! can be met with 13 .txf7+! ?, while after 1 2 . . . 'ifxd 1+?! 13 l:txd l Black has trouble completing his development, but the exchange sacrifice 12 . . . l'Llc6 1 3 .txf8 'ifxf8 promises him sufficient counterchances.

Considering that the white king lacks a secure shelter, Black's desire to keep the queens on the board is understandable in general terms, but the text-move has a dubious reputation for a very concrete reason.

12 'ifd2?! 12 .txf7+! (D) is considered to give White a

big advantage.

B

The lines after 1 2 . . . l:txf7 1 3 e6 l:tf8 14 'ifb3 l'Lla6 15 e7+ l:tf7 16 'ifxb7 l:te8 are very sharp. However, White's decision to avoid these favourable complications is not objectively justified. The move in the game does not aspire to an advantage; on the contrary, the initiative passes to Black.

The other aggressive continuation is less effective: 1 2 e6?! fxe6 1 3 .txe6+ 'iii>h8 opens yet another centre file and gives Black good counterplay.

12 ••• t'Llc6 13 f4 Black can meet the attempt to take advantage

of his queen's precarious position by 1 3 h3 l'Llxe5 14 .te2 with the pawn sacrifice 14 . . . h6 1 5 .txh6 .txh6 16 'ifxh6 c6; the dark-squared blockade is more significant than the slight material loss.

13 ••• l:lad8 14 l'Lld5 l'Llxe5 In view of the white king's position, Black's

desire to clarify the situation in the centre to the maximum is understandable. The flanking blow 14 . . . g5 is also interesting.

15 fxe5 c6 (D) 16 .td3? White refuses to acknowledge what has hap

pened. Winning the exchange this way or after 1 6 0-0? 'ifg4+ 17 'iii>h 1 'ifxe4+ 1 8 f3 'ifxe5 19 .tc5 cxd5 20 .txf8 puts him on the brink of defeat, perhaps even beyond it. It was time to

LINES WITH iJ..J4 OR iJ..g5 115

w

begin the struggle for salvation with 1 6 'ilke2 cxd5 17 .txd5 lLlf4 1 8 .txf4 'ilkxf4 1 9 'ilke3 (or 1 9 0-0), when the powerful bishop on d5 grants White good chances of survival.

16 ••• cxd5 17 .tg5 'ilkg4 18 .txd8 l:txd8 19 f3 White sobers up. Black has many ways of cre

ating threats, and it is too late to try to stabilize the position, so White starts seeking counterplay without worrying too much about material parity. The pawn is given up in order to open the ffile and target f7.

19 .•. 'ilkxf3 20 l:tn 'ilkh3 21 'ilkg5 l:te8 22 .te2 .i.xe5 23 .txh5 .txb2?! (D)

The point of White's operation is that the endgame after 23 . . . 'ilkxh5? 24 'ilkxh5 gxh5 25 l:td1 provides real chances of salvation. However, if Black inserts 23 . . . h6! he wins quickly. The text-move prolongs the fight.

w

24 .te2 .txcl 25 'ilkxcl 'ilkxh2 The choice of the pawn to capture - this one

or the one on e4 (after 25 . . . 'ilkh4+) - is a matter of taste, but the remaining centre pawn to a certain extent justifies White's search for counterplay.

26 .i.d3 'ii'xa2

After 26 . .. 'ilk g3+ 27 �e2 f5 the realization of Black's advantage is easier.

27 'ilkc7 'ilk at+ 28 �d2 'ilkb2+ 29 �dl :rs 30 exd5 'ilkb3+ 31 �e2 .l:.e8+ 32 �d2 'ilkxd5 33 .l:.xf7 (D)

B

With the rooks off the board, the white king does not have to worry about mating threats. Four pawns outrank the bishop, of course, but the technical stage is not simple. First, pawn advances expose the black king and increase White's chances of chasing it; second, the queen and the bishop combo is capable of stopping either passed pawn pair on its own, so the pairs' alternating motion must be coordinated.

33 .•• 'ilkxf7 34 .tc4 .l:.e6 35 'ilkc8+ �g7 36 .txe6 'ilkf4+ 37 �e2 'ilke4+ 38 �d2 �f6 39 .tc4 'ilkd4+ 40 .td3 'iVb4+ 41 �e2 'iVe7+ 42 �f3 �g7 43 .tc4?

This allows an exchange of queens that solves all of Black's technical problems.

43 ... 'ilkf8+ 44 'iVxfS+ �xf8 45 �f4 aS 46 .tb5 �e7 47 �es h5 48 .td3 g5 49 .te2 h4 50 .tg4 a4 51 �d4 �d6 0-1

Let's sum up the results. The variations examined in this chapter do not aspire to refute our opening and shouldn't cause Black to panic. At the same time they are based on the fundamental principles governing the initial stage of the game and are quite acceptable for the purpose of fighting for an advantage. The basis of these variations was laid down long ago, and there aren't many notable innovations, although certain changes in the popularity and the assessment of this or that continuation do take place. In particular, today the system with 4 .i.g5 is in vogue.

8 Closed Systems: White Plays e3

In this concluding chapter we shall examine systems that at first sight are the least ambitious for White. He does not strive to seize the centre immediately, and nor does he seek an active development of his dark-squared bishop in the early stages of the game.

However, this approach has its positive side too. Firstly, omitting a move by the bishop saves a tempo that will be useful in developing the initiative in some other direction; second, the bishop is ready to nip in the bud any counterplay along the a5-el diagonal or, following its opposing number's departure for the long diagonal, seek gainful employment on another one - a3-f8.

While logic informs us that the battle for the maximal achievements has to be waged in the main theatres of war, which are covered here in Chapters 3-6, the variations cited in this chapter give White some scope for optimism. After all, he is a tempo up on his opponent, and familiarity with the subtleties and experience of playing the resulting positions can make these lines into fairly potent weapons in practice.

In the lines of Game 24 (B.Socko-Gopal) White strives for active play on the queenside, in the main game with b4, while there are also lines where the queen moves to b3 in the style of the Russian System, but with the move e3 played first.

In Game 25 (Tregubov-Jaracz), we examine lines in which White develops his king's bishop first of all, in the main game by exchanging on d5 and then playing ..i.c4. The other options covered feature the moves �e2 and .id3; in all these lines, Black will generally play . . . c5 at some stage, though a solid approach with .. c6 is feasible too, given that White's own play is a little sluggish.

Game 24 [094]

Bartosz Socko - Geetha Gopal Gibraltar 2008

1 d4 tt::lf6 2 c4 g6 3 tt::lc3 d5 4 tt::lf3 Is there any point in saving time on the de

velopment of the knight and playing 4 e3 first? In most cases this is a mere transposition, with the only notable independent line being perhaps 4 . . . �g7 5 'iVb3 e6 6 'ifa3, for the moment holding up Black's castling. Then 6 . . . 'ife7 7 'ifxe7+ �xe7 cannot be recommended because of 8 b3, when the bishop continues to harass the black king from a3, while its exchange would leave the dark squares weak. In response to 6 . . . �f8 White does not have to go back with a repetition but can try 7 'ifa4+. Black has several ways to solve the resulting problems, such as with the . . . tt::lc6-e7 manoeuvre.

4 ••• �g7 5 e3 0-0 (D)

w

This is the starting position of this chapter's systems. Having left the c 1 -bishop on its initial square, White reckons that the tempo he saved will prove useful in seizing the initiative in

CLOSED SYSTEMS: WHITE PLAYS e3 117

another development set-up. He has several options to this end.

6 b4 Before discussing the ramifications of this

queenside space-grab, we shall cover several of White's other ideas.

Since we have already mentioned the development of the queen to b3, let's follow this direction further - 6 "ifb3 (D).

B

The principal difference between this version and the games in Chapter 6 is that White is ready to recapture on c4 with the bishop and is not obliged to centralize his queen excessively. Black has two plans:

I ) Maintaining the pawn on d5 (lines 'a' and 'b'). This can be achieved by bolstering it with one of its neighbours on the left or on the right. Why do the continuations that we previously considered dubious tum out to be acceptable in this case? The explanation lies in the position of White's dark-squared bishop - confined by its own pawn-chain, it does not participate in taking advantage of the weak squares that now appear in Black's camp.

2) Ceding the centre with the hope of exposing the drawbacks of the set-up with the queen on b3 and the bishop on c4 (line 'c').

Having set the scene, let's tum to the specifics:

a) The position after 6 . . . c6 can be claimed with equal justification by both the Griinfeld Defence and the Slav. Since by most of the major classifications it formally belongs to our opening, we shall devote some attention to it. Besides reinforcing the centre, Black also does not rule out . . . dxc4 followed by the standard advance of the queenside pawns. Here 7 .id2 looks logical - first, White maintains the so-called

"struggle for a tempo"; in case of the capture on c4 he will recapture with the bishop from fl in one leap; second, he intends to deploy his forces with .l:.c l , cxd5 and lLlb5. This is frequently met with 7 . . . e6, securely defending the central bastion and in case of an exchange on d5 planning to recapture with the e-pawn. However, that shelves the c8-bishop's hopes for an honourable future, and although Black has no terminal weaknesses, he will have to fight in cramped quarters for the next phase of the struggle.

b) If the d5-pawn's other neighbour steps forward, i.e. 6 . . . e6, then subsequent play resembles the Tartakower Queen's Gambit: the queen 's bishop is fianchettoed on b7, followed by its eventual activation. Here is an example: 7 .id2 b6 8 .ie2 .ib7 9 0-0 lLlbd7 10 cxd5 lLlxd5 1 1 lLla4 e5, Burmakin-Navara, Dubai 2005.

c) Naturally, the second plan is initiated with 6 . . . dxc4, and after 7 .ixc4 (D) the path forks:

B

c 1) There is the typical undermining thrust 7 . . . c5, when the natural reaction 8 d5, restricting the scope of the queen-and-bishop tandem, doesn't go too well with the preceding play. The other continuation is 8 dxc5 "ifa5, and the pattern of subsequent play is already familiar to us.

c2) 7 ... lLlc6 threatens ... lLla5 but allows complications after 8 lLle5. The exchange 8 . . . lLlxe5 9 dxe5 is forced. Now 9 . . . lt:Jg4 10 f4 restricts the black knight's rights, while 9 . . . lt:Jd7 leads to a sharp and unclear game after 10 e6. The other continuation 1 0 i.xf7+ ? ! is risky: 10 . . . �h8 1 1 .ic4 ( 1 1 f4? loses a piece to l l . . .e6) l l . . .lLlxe5, and White has wasted time on an exchanging operation.


Let's move on to the next attempt to fight for the initiative on the queenside - 6 �d2 (D).

B

In contrast to the similar manoeuvre in Game 6, White is not planning to carry out the e4 advance, as the loss of tempo renders this entirely toothless. Rather, he makes a useful move and invites Black to clarify his intentions, which will determine the development of the other white pieces. Black has many options, the most principled being 6 . . . cS, considering that the bishop has temporarily obstructed the white queen's control over the centre squares. A temporary pawn sacrifice secures Black rapid and comfortable development: 7 dxcS lt:Ja6 8 cxdS lt:JxcS 9 �c4 �fS 10 0-0 .l:lc8 1 1 �e2 lt:Jfe4, and Black can count at the very least on restoring the material equality.

We now return to 6 b4 (D):

B

A blunt move, at once commencing a queenside attack. As is the case with all pawn advances, it leaves in its wake weaknesses that are not significant should the initiative succeed, but will be sensitive in case of its failure. Pawn moves in the opening also use up tempi at the expense of piece development, and constant

care should be taken that the lag in development does not assume dangerous proportions.

6 •.. c6 It would appear that Black can forget all

about . . . cS, but nevertheless this undermining counterblow is still possible - in two versions, even:

a) After the introductory 6 . . . a5 7 bS cS all the same 8 bxc6 bxc6 9 �a3 restores White's control over the cS-square, whereas the aSpawn takes an important square away from its own queen.

b) 6 . . . b6 clearly prepares . . . cS. Then: b l ) The main continuation used to be 7

�b3, when Black can treat the position in the spirit of the Tartakower Queen's Gambit with 7 . . . �b7 or choose the thematic 7 . . . cS .

b2) Recently 7 �a3 cS 8 bxcS bxcS (D) has become popular.

w

After 9 .l:lcl (the immediate 9 �xeS? gives Black a serious initiative after 9 . . . lt:Ja6 10 �a3 �aS 1 1 �b2 .l:lb8) 9 . . . lt:Je4 White can choose between playing for piece activity, or the pawngrab under improved circumstances:

b2 1 ) 1 0 cxdS lt:Jxc3 1 1 .l:lxc3 cxd4 1 2 lt:Jxd4 �xdS 1 3 �b3 �e6 14 �xdS �xdS l S �bS a6 16 �a4 .l:la7 and White's activity was neutralized in Korchnoi-Svidler, European Clubs Cup, Fiigen 2006.

b22) 10 �xeS ! ? poses interesting problems; e.g., l O . . . lt:JxcS ( l O . . . �aS 1 1 �a4) 1 1 dxcS �aS 12 �d2 lt:Ja6 ( 1 2 . . . dxc4 1 3 �xc4 lt:Jc6 14 lt:JdS) 1 3 c6 .l:ld8 14 lt:JxdS �a3 1S lt:Jf6+ �xf6 16 �xd8+ �g7 1 7 �d2 .l:lb8 1 8 lt:Jd4 .l:lb2 (Bilobrk-Berkes, Bosnian Team Ch, Vogosca 2007) 1 9 �c3 'ir'xa2 20 cS .l:lb1 2 1 �c4 and White is consolidating.

We now return to 6 . . . c6 (D):


w

This ushers in another plan of counterplay aiming to expose the minuses of the early b4. Reinforcing the centre, Black retains the right to hobble the b4-pawn, followed by its exchange for his a-pawn; the c-pawn can be used then to clear the long diagonal.

7 .i.b2 .i.fS This is one of several natural continuations:

prior to commencing the planned operation, Black makes a useful move, waiting for the fl bishop to leave its place.

Also played are the immediate 7 . . . ltJe4 and 7 . . . .i.g4, and the attempt to initiate at once concrete play with 7 . . . a5.

8 .i.e2 dxc4 In this system this general idea has been

known for a long time, but at precisely this point it is relatively new. Usually 8 . . . l0e4 or 8 . . . l0bd7 has been played here.

9 .i.xc4 bS 10 .i.b3 aS (D)

w

Since maintaining a pawn on b4 with I I a3 is not prudent - it will prove to be weaker than Black's on c6 - the black queenside pawns acquire mobility and are about to raise the question of who has the initiative in this sector of the board.

11 bxaS 'ii'xaS 12 0-0 l0bd7 13 'ii'e2 By attacking the b5-pawn, White forestalls

the . . . c5 advance, while also threatening to set his centre in motion.

13 ••• b4 Thus Black drives the knight away. 14 l0a4 Underestimating the reply; White should

have retreated to bl with a roughly even game. 14 ••• 'i!r'bS (D)

w

Black is ready to devalue his pawn-island in order to give his opponent a weakness on c5 and, crucially, seize the powerful outpost on c3.

1S 'ii'xbS 15 'i!r'd I is hardly any better, as then the

queen could face an unpleasant opposition from a rook on d8, and it would still be difficult to bring the a4-knight back into the game.

1S ••• cxbS 16 tOeS ltJxcS 17 dxcS ltJe4 18 .i.xg7 ..ti>xg7 (D)

w

19 :rc1 The desire to get rid of the arch-nemesis -

the black knight - by means of 1 9 .i.d5 would deprive the a2-pawn of protection. Black would then gang up on it by doubling rooks, and


White could not use the b ! -square to counterattack along the b-file.

19 ••• ttJc3 20 �n l:taJ 21 ttJd4 The attempt to block the enemy bishop's di

agonal with the subsequent attack on the b4-pawn fails to solve White's problems: 21 ltJel l:tc8 22 f3 :xeS 23 e4 .i.e6 24 .i.xe6 (24 ltJc2 ltJxa2) 24 . . . fxe6 25 tiJd3 :c4 26 ltJe5 .l:tc8 27 tiJd3 �f6 28 ltJxb4 .l:tc4 29 tiJd3 ltJxe4.

21 .•• .l:.c8 22 tiJxfS+ gxfS 23 :c2 .!:.xeS 24 .l:.d1 ltJe4 25 .l:.b2 .l:.c3 (D)

w

26 g4? White's position cannot be envied, but this

move changes its evaluation from bad to lost.

Besides yet another pawn, Black obtains mating threats. Let's examine White's other options:

a) 26 .l:.d5? loses at once: 26 . . . .l:.c 1 + 27 �e2 ltJc3+ 28 �d2 .l:txb3.

b) 26 .i.d5 ltJc5 27 .l:.xb4 e6 28 .i.f3 .l:.xa2 leads to a near-zugzwang situation. Now in order to secure the second rank against an attack, one rook must keep an eye on the c5-knight, and the other on the d3-square. Also, the bishop and the king must provide for the .i.e2 defence in reply to the . . . tiJd3 leap. Despite the fact that the battlefield has shrunk considerably, it is not clear what White is to do once he runs out of pawn moves.

c) White has a choice of rook endings a pawn down, such as 26 :d4 ltJc5 27 h4 ltJxb3 28 axb3 .l:.axb3 29 .l:.xb4 :xb4 30 .l:.xb4 .l:.c5.

d) There is the passive stance 26 .i.c2 tiJd6 27 .i.b3 ltJc4 28 .l:.bb 1 , inviting Black to demonstrate how he intends to strengthen his position.

26 ••• fxg4 27 .l:.d4 rs 28 .l:.xb4 :cl + 29 �g2 ltJgS 30 .i.dS .l:.d3 31 .l:.xbS f4 32 f3?!

This loses at once. White could fish for chances with 32 h4 gxh3+ 33 �h2 fxe3 34 fxe3 .l:.xe3 35 a4.

32 ••• ltJh3 33 fxg4 fxe3 34 .l:.Sb3 lDf4+ 35 �g3 .l:.xb3 36 .l:.xb3 ltJxdS 0-1

Game 25 [094]

Pavel Tregubov - Pawel Jaracz European Ch, Warsaw 2005

1 d4 tiJf6 2 c4 g6 3 ttJc3 dS 4 tiJf3 .i.g7 5 e3 In the previous game we devoted attention to

lines with 'i&'b3 and b4. Our focus here will be on w lines where White adopts a very natural approach - first of all he develops his light-squared bishop, either to e2 or d3, or (as in the main game) to c4 after an exchange on d5.

5 ••• 0-0 Although for a true Griinfeld fan the c-pawn's

road to glory leads to c5, here the self-imposed restriction of White's dark-squared bishop adds arguments in favour of 5 . . . c6 (D).

This position is very popular, and is reached more commonly via the initial moves of the Slav Defence - today's Closed opening No. I . Black maintains pawn parity i n the centre,

which conforms to the classical principles of the game in the opening stage.

Of course, having chosen . . . c6, Black will rarely profit from playing a quick . . . c5, given


the loss of tempo. Therefore Black builds a pawn-wedge on the light squares - c6-d5-e6 -generally after first exchanging his bishop on f3. He has no weaknesses or bad pieces, but having ceded the bishop-pair, he should maintain the closed character of the fight, which diminishes his hopes for active counterplay. A protracted struggle, where White strives to obtain space for his bishops, is likely. One sample line is 6 �e2 0-0 7 0-0 a6 8 b3 ltJe4 9 �b2 ltJxc3 10 �xc3 �g4 1 1 h3 �xf3 1 2 �xf3 e6, Carlsen-Bu Xiangzhi, Biel 2007. However, we shall not be going into further detail on this line, on which there is extensive theory due to its importance to the Slav Defence, as there is no need for a Griinfeld player to restrict himself to the . . . c6 approach.

6 cxd5 In the previous game we covered the plans

that give priority seizing the initiative on the queenside. Here we shall examine lines where the kingside is developed more rapidly.

First we should note that 6 �d3 does not entice serious players, as the bishop obstructs the queen's file and the advance e4 will never come to pass, as Black at once replies 6 . . . c5, with a good game.

The most natural continuation is therefore 6 �e2 (D).

B

The most principled reply here is also 6 . . . c5. Then:

a) The typical reaction 7 dxc5 promises no advantage. After 7 .. .'�Wa5 8 0-0 (or 8 �d2) 8 . . . dxc4 9 �xc4 'iVxc5 the bishop's absence from f4 noticeably improves Black's mood, while following 8 cxd5 ltJxd5 9 'iVxd5 �xc3+ 10 �d2 l:!.d8 the queen sacrifice 1 1 'iVxd8+?! 'iVxd8 1 2 ixc3 is not completely sound, so

White would be prudent to agree to an equal ending by 1 1 �xc3 'iVxc3+ 1 2 bxc3 l:!.xd5.

b) Nowadays 7 0-0 cxd4 8 exd4 is preferred, transposing into positions that frequently arise from other openings. If within the next few moves Black exchanges on c4, then a standard IQP set-up appears on the board. If on the other hand Black is unwilling to determine his plans early on, White can go for the c5 advance, with the play typical of certain lines in the NimzoIndian or Caro-Kann.

Alternatively, Black can content himself with the fact that the fl -bishop has wasted a tempo, and exchange by 6 . . . dxc4 7 �xc4, and only then carry out 7 . . . c5. This position is known from the Queen's Gambit Accepted, with White already castled. In our case Black's extra move cannot hurt him, but nevertheless . . . c5 without exchanging first looks more logical, as it is hard to see how Black can benefit from the fact that the d4-pawn has acquired the ability to advance.

6 ••• ltJxd5 7 �c4 (D)

B

Thus White brings out his bishop to c4 in one move, while at the same time the black knight also changes its position. It only remains to find out if this favours either side in some way.

7 ... ltJxc3 There is no sensible way to keep the knight

on d5, so the other option is 7 . . . ltJb6. After the obvious 8 �b3 (D) Black has to determine with which pawn to undermine his opponent's chain:

a) If this mission is assigned to the e-pawn, then in order to prepare it, the queen's knight has to come out to c6 or d7. Naturally, 8 . . . ltJc6 is the more active. After 9 0-0 e5 a double capture on e5 with the aim of obtaining a mobile e- and f-pawn phalanx lacks support due to

122 CHESS EXPI.AINED: THE GRONFELD

8

insufficient development, and the play is centred around a formation that we have seen several times after 10 d5 ltJa5, followed by the blockade of White's centre or its remnant - the pawn on d5.

b) The undermining thrust from the other side can be carried out without preparations: 8 ... c5 . The attempt to hold on to the material advantage with 9 dxc5 �xd 1 + 10 �xd I ltJ6d7 I I ltJd5 ltJc6 1 2 ltJd4 (or 1 2 ..ta4) has failed to bring White success, as Black obtains sufficient compensation. White usually chooses 9 0-0 (D).

B

Now: bl) With 9 . . . c4, Black demonstrated an origi

nal treatment in Pelletier-Mamedyarov, European Clubs Cup, Kerner 2007. Such an advance is often considered dubious, because White's central majority ranks higher, but in this case Black's active pieces prevent the advance of the centre pawn group: I 0 ..tc2 ltJc6 I I �e2 ..tg4 1 2 h3 ..txf3 1 3 �xf3 e5.

b2) After 9 . . . cxd4 10 exd4 ltJc6 1 1 d5 ltJa5 12 l:tel (D) the standard pawn-formation appears on the board, with the slightly novel position of the bishop on b3.

B

Here we shall note two dashing attacks by White, the first one being successful, though the overall assessment is unclear:

b21 ) 1 2 . . . l:te8 1 3 ltJe5 ltJd7 14 ltJxf7 should lead to equality, but Black went awry in Tregubov-Macieja, Bundesliga 2006/7.

b22) Events developed along a less fortunate course for White after 1 2 . . . ..tg4 1 3 ..tg5 .l:.e8 14 d6 �xd6 1 5 ..txf7+ �xf7 1 6 l:.xe7+ �xe7 17 ..txe7 �xe7 in Tregubov-Kumosov, Moscow 2008. In the ensuing play White failed to take advantage of the enemy king's position in the centre, and Black's compensation turned out to be more important.

8 bxc3 (D)

B

Similar pawn-formations can be seen in other openings as well, including the Queen's Gambit and the Queen's Indian Defence. It is also typical of the Griinfeld Defence, the difference being the superior placement of the bishop on g7. A significant portion of this book is devoted to similar positions with the difference that White's e-pawn has advanced to e4. Nothing prevents it from stepping forward here as well, but for the moment that would not be to White's advantage: first, a tempo is lost, and second,


holding the centre is considered to be more difficult with the bishop on c4 and the knight on f3, and Black's counterplay has good chances of success. Once White completes his development, though, the advance of the centre pawns will be on the agenda, provided Black doesn't force him to solve other problems in the meantime. For now the c3-d4-e3 wedge fulfils its task, restricting Black' s dark-squared bishop, although it is hard to call its opposing white number active either, and that is one of White's main concerns.

Let's list Black's next few moves - the cpawn goes to c5, the knight to a5, the queen's bishop to b7 or g4. His plans include setting up a blockade on the light squares, operations along the c-file, and the advance . . . e5 is also not ruled out. The choice of the actual move-order and prioritizing the plans is up to the individual player.

8 .. .lbc6 It is more common to begin with 8 ... c5, when

a topical position is reached after 9 0-0 "file? I 0 "flle2. White gears up for the e4 advance, if necessary agreeing to station the queen's bishop on the unappealing post on b2 for the time being, in order to bolster the d4-pawn.

9 0-0 lL:la5 10 i.e2 b6 White retreated the bishop to e2 so as not to

obstruct the d-file, but this deprived the advance of the e-pawn of support. Black therefore decides that a fianchetto of his queen's bishop is more expedient in this case.

11 i.a3 i.b7 12 "fllc2 c5 (D)

w

With the plan of development chosen by White, his opponent's desire to open the c-file is natural.

13 dxc5

If White allows an exchange on d4, the major pieces will start disappearing on the c-file, and there are no apparent reasons to expect an advantage in the minor-piece ending. Therefore White attempts to obtain the initiative on the queenside, counting on the newly-opened files, the a5-knight's insecure position and the weakness of the enemy pawn on c5. The move played does not aim at winning a pawn, as a further capture on b6 would activate Black's pieces with the likely elimination of all the pawns on the queenside; on the contrary, White is striving to close the c-file.

13 •.• "fl/c7 14 llab1 llfc8 (D)

w

15 .l:tfcl In order to force Black to capture on c5,

White prepares to push his pawn to c4; then the exchange on b6 will be a real threat, as it will be followed by the c5 advance.

15 ••• e6 This opens another line of work for the dark

squared bishop but weakens the long diagonal. At the moment it is not clear if this has any significance.

16 c4 bxc5 17 .l:tb5 (D) The ensuing forced play is tempting but not

mandatory. 17 i.b2, offering to exchange bishops, was not a bad idea.

17 ••• i.c6 If Black defends the pawn by 17 . . . i.f8, White

gets the same attack with 1 8 "fllc3 and 1 9 i.b2 for free.

18 .l:txc5 i.f8 After 1 8 . . . lL:lb7 1 9 .l:tg5 the rook is out of ac

tion and has to find a way home, although that does not fully compensate for the loss of a pawn. If Black blocks its way with 1 8 . . . e5, White will get activity and two pawns for the


B

exchange: 19 �c3 �b7 20 �xc6 �xc6 2 1 �xeS and 22 f4.

19 �c3 �xeS 20 �b2 eS An attempt to flee with the king by 20 . . .'�f8?

fails: 21 �h8+ �e7 22 �f6+ �d7 23 l:tdl + �d5 (23 . . . .i.d6 24 �e5#) 24 'ili'xh7, and the extra rook is of no help to such a king.

21 �xe5 l:tab8 (D) Taking advantage of the momentary respite,

Black intends to destroy the terrible battery that is lined up on the long diagonal. This is also the only defence; after 2 l . . . f6? White has a pleasant choice between 22 �xc6 �xc6 23 �xa5 and 22 �g4.

w

22 �d7

Now the game ends in a draw. Preserving the bishop is a tempting idea, but after 22 �al ? ! �b4 23 'ifd4 'ifd6 24 �xc6 'ili'xd4 25 �xd4 �a3 26 l:tdl �xc4 White's compensation is already insufficient. 22 �xg6 leads to a position that is hard to evaluate after 22 . . . fxg6 23 'ifh8+ �f7 24 �xh7+ �e8 25 'i!i'xg6+ �f7 26 �g5. If White wants to continue the fight, the line 22 �g4 l:txb2 23 �h6+ �f8 24 'ifxb2 is available.

22 ••• l:txb2 23 �f6+ �f8 24 �xh7+ �g8 (D)

w

Going into the centre would be imprudent, as after the capture on b2 White obtains three pawns for the piece and the possibility of creating new threats. If White takes the rook now, the mechanism of the perpetual check works for a while, but White is unlikely to create any sort of serious attack with just two pieces.

25 �f6+ �f8 26 �h7+ �g8 27 �f6+ �f8 28 �h7+ lfz.llz

We can note the current trend of growing popularity of such restrained play in the opening. This can be explained in part by White's weariness of the main continuations that demand progressively more extensive opening knowledge and in part by the desire to play more solidly and to avoid rocking the boat excessively.

List of Games

Levon Aronian - Alexei Shirov, Candidates match (game 5 ), Elista 2007 43 Vlastimil Babula - David Navara, Czech Ch, Luhacovice 2003 88 Antony Bellaiche - Emil Sutovsky, Catalan Bay 2007 84 Viacheslav Eingorn - Vladimir Malaniuk, Baku 1979 62 Boris Gelfand - Alexei Shirov, Odessa (rapid) 2007 76 Humpy Koneru - Alexander Areshchenko, Lausanne 2006 20 Mikhail Kobaliya - Alexander Areshchenko, isle of Man 2006 48 Yuri Kuzubov - Pavel Elianov, Moscow 2005 I 03 Alexander Lastin - Valeri Yandemirov, Voronezh 2007 28 Viktor Laznicka - Ruslan Ponomariov, Karlovy Vary 2007 24 Jifi Lechtynsky - Jii'i Stocek, Czech Ch, Karlovy Vary 2005 I l l Vladimir Malakhov - Shakhriyar Mamedyarov, Spanish Team Ch, Sant Lluis 2005 6 Hikaru Nakamura - Jordi Fluvia Poyatos, Barcelona 2007 98 Hikaru Nakamura - Sergey Kudrin, USA Ch, San Diego 2004 1 07 Evgeny Nayer - lbragim Khamrakulov, Ubeda 2001 52 Evgeny Nayer - Igor Kurnosov, Russian Ch, K rasnoiarsk 2007 93 Peter Heine Nielsen - Magnus Carlsen, Bundesliga 200415 15 Tomi Nyback - Michal Krasenkow, European Ch, Antalya 2004 39 Zoltan Ribli - Andrei Volokitin, Slovenian Team Ch, Celje 2004 I 0 Alexander Shabalov - Alexander Areshchenko, Port Erin 2005 69 Bartosz Socko - Geetha Gopal, Gibraltar 2008 I I6 Pierre Yves Toulzac - Valentin Bogdanov, Malakoff2007 32 Pavel Tregubov - Pawel Jaracz, European Ch, Warsaw 2005 120 Anatoli Valsser - Maxime Vachier-Lagrave, French Ch, Chartres 2005 65 Nikita Vitiugov - Zhao Jun, World Junior Ch, Erevan 2006 56

I ndex of Variations

A : 1 d4 ltJf6 2 c4 g6 3 ltJc3 d5 4 cxd5 B: 1 d4 ltJf6 2 c4 g6 3 ltJc3 d5 without

4 cxd5 C: Other Lines

A)

w

5 ..td2 28 5 ltJa4 32

1 d4 ltJf6 2 c4 g6 3 ltJc3 d5

4 cxd5 ltJxd5 5 e4

5 ltJxc3 5 . . . ltJb6?! 24

6 bxc3 ..tg7 39

w

7 ltJf3 52 7 ..te3 52

7 ..tc4 39 7 . . . c5 8 ltJe2 ltJc6 9 ..te3 0-0 10 0-0 41: a) 10 . . . ltJa5 41 b) 10 . . . 1i'c7 48 c) 1 0 . . . ..td7 48 d) 10 . . . cxd4 1 1 cxd4 ..tg4 44 e) 10 . . . ..tg4 43

7 c5

w

8 l:tb1 61 8 ..te2 52 8 ..te3 56

8 0-0 8 . . . 1i'a5 65 8 . . . cxd4 9 cxd4 'ii'a5+ 65 8 . . . ltJc6 66

9 ..te2 61 9 cxd4

9 . . . b6 62 9 . . . 1i'a5 76 9 . . . ..tg4 66 9 . . . ltJc6 66 10 d5 : a) I O . . . ..txc3+ 66 b) IO . . . ltJe5 69

10 cxd4 'ii'a5+ 77

B)

1 d4 ltJf6 2 c4 g6 3 ltJc3 d5 4 ltJf3

4 e3 116

INDEX OF VARIATIONS 127

4 'ii'b3 84 4 'ii'a4+ 84 4 ..ig5 98 4 ..if4 107 4 . . . ..tg7: a) 5 e3 107 b) 5 l:k l 107 c) 5 li:Jf3 1 J J

4

w

5 'ii'b3 83 a) 5 'ii'a4+ 84 b) 5 ..ig5 103 c) 5 ..if4 111 d) 5 cxd5 li:Jxd5: d l ) 6 li:Ja4 33 d2) 6 e4 li:Jxc3 7 bxc3 52

..ig7

e) 5 e3 116 5 . . . 0-0 (5 . . . c6 120): e l ) 6 b4 1 1 7 e2) 6 'ii'b3 J / 7 e3) 6 ..id2 1 18 e4) 6 ..ie2 121 e5) 6 cxd5 li:Jxd5 7 ..ic4 121

5 . . . c6 84

Now: 7 . . . a6 85 7 . . . ..ig4 88 7 . . . li:Jc6 88 7 . . . li:Ja6 93

C)

5 dxc4

6 'ii'xc4 0-0 85 7 e4 85

1 d4 1 c4 li:Jf6 2 li:Jf3 g6 3 g3 ..ig7 4 ..ig2: a) 4 . . . c6 5 0-0 0-0 6 d4 d5 I I b) 4 . . . 0-0 5 0-0 d5 6 cxd5 li:Jxd5 7 li:Jc3 li:Jc6 8 d4 tt:lb6 16

2 li:Jf3 g6:

1 li:Jf6 2 c4

a) 3 c4 - 2 c4 g6 3 li:Jf3 b) 3 g3 ..ig7 4 .i.g2 d5 5 0-0: b l ) 5 . . . 0-0 6 c4 - 2 c4 g6 3 g3 ..ig7 4 ..ig2 d5 5 li:Jj3 0-0 6 0-0 b2) 5 . . . c6: b21 ) 6 li:Jbd2 0-0 7 c4 10 b22) 6 c4: 6 . . . dxc4 10; 6 . . . 0-0 I I

2 g6 3 g3

3 f3 d5 4 cxd5 li:Jxd5 5 e4 li:Jb6 24 3 li:Jf3 ..ig7 (3 . . . d5? ! 25) 4 li:Jc3 d5 - see B

3 ... ..ig7 3 . . . c6 4 ..ig2 d5 5 cxd5 cxd5 6 li:Jf3 ..ig7 7 0-0 0-0 7

4 ..ig2 d5

w

5 li:Jf3 5 cxd5 li:Jxd5 6 e4 20

5 0-0 5 . . . c6: a) 6 0-0: 6 ... dxc4 10; 6 . . . 0-0 1 I b) 6 cxd5 cxd5 7 0-0 0-0 7

6 0-0 6 cxd5 li:Jxd5 7 0-0: a) 7 .. . c5 16 b) 7 . . . li:Jb6 8 li:Jc3 li:Jc6 16

6 c6 1 1 6 . . . dxc4 15

Or: 7 li:Jc3 11 7 b3 11 7 li:Je5 11 7 'ii'b3 11 7 li:Jbd2 10

7 cxd5

7 cxd5 7

Chess Explained: The Main-Line Slav David Vigorito This book discusses all major lines after 1 d4 d5 2 c4 c6 3 lt:Jf3 lt::lf6 4 lt::lc3 dxc4. The lack of symmetry provides scope for creative players to obtain winning chances with either colour. Play can take many strategic directions: sometimes the centre is blasted open and piece-play dominates, while in other cases the centre becomes fixed, or remains dynamic well into the middlegame. 1 12 pages, 248 x 1 72 mm; $24.95 I £12. 99

Chess Explained: The Queen's Gambit Declined James Rizzitano The Queen's Gambit Declined is one of the most important and popular of chess openings - an immensely solid and classical choice for Black that has remained topical for more than a century. 128 pages, 248 x 1 72 mm; $23.95 I £12.99

Chess Explained: The English Opening Zenon Franco The English Opening is a flexible and dynamic choice for White, which avoids a great deal of sharp and well-mapped opening theory. It is popular with all levels of chess-players, and has been used to good effect at world championship level by Kasparov, Korchnoi, Botvinnik and other greats of the game. 112 pages, 248 x 1 72 mm; $19. 95 I £12. 99

Chess Explained: The Classical Sicilian Alex Yermolinsky Fonner US Champion Yennolinsky explains, in masterful fashion, the key plans and ideas of one of his favourite openings. This dynamic Sicilian weapon has been a favourite of players such as Anand, Kramnik and Shirov. 112 pages, 248 x 1 72 mm; $19.95 1 £12.99

Chess Explained: The c3 Sicilian Sam Collins The c3 Sicilian is a popular way to sidestep the main lines of the Sicilian. White denies his opponent the chance to demonstrate his preparation in some heavily-analysed system; instead Black will have to fight carefully for equality in a more rational position of White's choosing. 1 12 pages, 248 x 1 72 mm; $19. 95 I £12.99

Chess Explained: The French Viacheslav Eingom & Valentin Bogdanov This important defence suits a wide range of chess temperaments - an opening that is solid yet uncompromising. In many critical variations White accepts major structural weaknesses in return for piece-play and dynamic chances. 128 pages, 248 x 1 72 mm; $23.951£12.99

Chess Explained: The Nimzo-l nd ian Reinaldo Vera Without doubt the Nimzo is one of the most important of all chess openings, and popular at all levels of play. It is considered a fully reliable defence, and provides winning chances for both sides as it leads to structures of great strategic variety and complexity. Key battlegrounds include the blockade, IQP positions, the handling of unbalanced pawn-structures, and the struggle between bishop and knight. 112 pages, 248 x 1 72 mm; $24.95 I £12.99

Chess Explained: The Queen's lndiarr Peter Wells 1 d4 lt::lf6 2 c4 e6 3 lt:Jf3 b6 is a mainstream strategic opening, rich in nuances. Many of the modem main lines involve moves that look extravagant, but are backed up by a deep underlying logic. 128 pages, 248 x 1 72 mm; $19.95 I £12.99

About the Publisher: Gambit chess opening books are designed to be different. We recruit the finest authors, writing passionately on their specialist systems. Our editorial team is obsessive about double-checking variations and transpositions. Gambit believes trustworthy analysis is important, and that an openings book should remain an asset over many years.

www.gambitbooks.com

[ijAI��I Bll T/ www.gambitbooks.com

Chess Explained books provide an understanding of an opening and the middlegames to which it leads, enabling you to find the right moves and plans in your own games. It is as if you were sitting at the board with a chess coach answering your questions about the plans for both sides. the ideas behind particular moves, and what specific knowledge you need to have.

The Grunfeld Defence is one of the most aggressive openings at Black's disposal. He creates immediate imbalance and plans to strike at White's centre with all available resources. An insipid response from White

may not even retain equality, so in the main . lines he takes up the gauntlet: he creates a large pawn-centre and puts this to use to launch an attack. While the theory of these lines has been extensively developed, there

is a coherent logical thread running through them: the Grunfeld is essentially an

opening of grand ideas, which need to be understood well in order to get to grips with the theory and handle the resulting positions. This book provides a discussion of all major lines, and a special contribution from Viacheslav Eingorn on the key ideas of the �b1 Exchange main line, which he was instrumental in developing in the 1970s and 1980s.

International Master Valentin Bogdanov has over thirty years' experience as a chess trainer, and is from Ukraine. His pupils include Moskalenko, Savchenko and Drozdovsky, and he has acted as a second for the well-known grandmaster and theoretician Viacheslav Eingorn since the late 1970s.

£12.99 $24.95

ISBN-13: 978-1-906454-06-7

ISBN-10: 1-906454-06-X

chess explained - ia600907.us.archive.org

Documents